884
#*approach privacy preserving data mining recent years privacy preserving data mining important problem large personal data business applications cases users provide personal information privacy sensitive information guaranteed paper propose new framework privacy preserving data mining multi dimensional data previous work privacy preserving data mining uses approach data distributions order perform mining approach dimension correlations di erent dimensions addition requires development new distribution based algorithm data mining problem does use multi dimensional records uses aggregate distributions data input leads fundamental design data mining algorithms paper develop new approach privacy preserving data mining does require new problem specific algorithms maps original data set new data set data closely matches characteristics original data including correlations di erent dimensions present empirical results ectiveness method 
#@Charu C. Aggarwal, Philip S. Yu
#t2009
#c0
#index0
#id0
#%1,2,3,4,5,6,7,8,9,10
#!

#*design privacy preserving data mining algorithms increasing ability track collect large amounts data use current hardware technology lead development data mining algorithms user privacy recently proposed technique addresses issue privacy data distributions aggregate level order perform mining method able privacy information implicit original attributes 
#@Dakshi Agrawal, Charu C. Aggarwal
#t2009
#c1
#index1
#id1
#%
#!

#*security privacy implications data mining data mining enables discover information security privacy issue make position paper discusses solutions ideas technology provides number mining using analyze data help 
#@Chris Clifton, Don Marks
#t2009
#c2
#index2
#id2
#%
#!

#*privacy preserving association rule mining partitioned data privacy considerations data mining paper addresses problem association rule mining transactions distributed sources site attributes transaction sites identify valid association rules sites reveal individual transaction data present party algorithm efficiently discovering frequent minimum support levels site individual transaction values categories subject descriptors database applications data mining systems distributed databases database protection 
#@Jaideep Vaidya, Chris Clifton
#t2009
#c3
#index3
#id3
#%1,11,12,13,14,15,16,17,18,19,20,21,22
#!

#*data balancing privacy precision mining logic rules recent data mining tools analysis large data little attention individual privacy issues introduce methods finding balance individuals right privacy data need general patterns detailed records particular focus data mining task classification decision trees base security control mechanism noise addition techniques used databases multidimensional matrix model statistical databases multidimensional line processing essentially noise addition techniques robust main noise addition techniques context statistical databases low statistical quality released statistics argue data mining major requirement security control mechanism addition privacy ensure precise free statistics high level knowledge constructed artificial data mining tools 
#@Vladimir Estivill-Castro, Ljiljana Brankovic
#t2009
#c4
#index4
#id4
#%
#!

#*privacy preserving mining association rules present framework mining association rules consisting categorical items randomized privacy individual actions feasible association rules privacy using uniform random discovered rules unfortunately privacy analyze nature propose class uniform derive estimator variance allow cover supports randomized datasets incorporate mining algorithms finally present experimental results applying real datasets 
#@Alexandre V. Evfimievski, Ramakrishnan Srikant, Rakesh Agrawal, Johannes Gehrke
#t2009
#c5
#index5
#id5
#%
#!

#*efficient approach clustering large multimedia databases noise clustering algorithms applied clustering large multimedia databases existing algorithms limited clustering multimedia databases requires cluster ing high dimensional feature vectors multimedia databases contain large amounts introduce new algorithm clustering large multimedia databases called density based clustering fi basic idea new approach model overall point density sum functions data clusters determining density clusters arbitrary shape easily described simple tion based overall density advantages new approach mathematical basis good clustering properties data sets large amounts noise allows compact mathematical clusters high dimensional data sets faster existing algo demonstrate ectiveness perform series experiments di erent data sets comparison shows new keywords clustering algorithms density based clustering high dimensional data clustering multimedia databases clustering presence noise introduction fast data stored databases increases types data stored come increasingly addition numerical data complex multimedia data age data stored cient retrieval com data usually transformed high dimensional feature examples feature vectors color histograms shape descriptors vectors text descriptors applications databases large millions data objects hundreds automated knowledge discovery large multimedia databases increasingly important research clustering trend detection databases di databases contain large amounts noise small portion large databases accounts addition known algorithms work high dimensional data methods 
#@Alexander Hinneburg, Daniel A. Keim
#t2009
#c3
#index6
#id6
#%
#!

#*data distortion probability distribution paper introduces data distortion probability distribution probability distortion involves steps step identify underlying density function original series estimate parameters density function second step generate series data estimated density function final step map generated series original data set probability distortion privacy individual original data set time probability series provides statistical properties original series distribution unlike conventional point distortion probability distortion difficult queries provides maximum statistical analysis 
#@Chong K. Liew, Uinam J. Choi, Chung J. Liew
#t2009
#c6
#index7
#id7
#%
#!

#*privacy interfaces information management facilitate sharing information using modern users able decide policy information reveal evolution user interfaces privacy policies sharing web experience shown privacy class objects policy 
#@Tessa A. Lau, Oren Etzioni, Daniel S. Weld
#t2009
#c7
#index8
#id8
#%
#!

#*maintaining data privacy association rule mining data mining services require accurate input data results meaningful privacy concerns influence users provide information investigate respect mining association rules users provide correct information mining process reasonable degree privacy present scheme based probabilistic distortion user data simultaneously provide high degree privacy user high level accuracy mining results performance scheme representative real synthetic datasets 
#@Shariq Rizvi, Jayant R. Haritsa
#t2009
#c8
#index9
#id9
#%
#!

#*privacy information generalization today great dissemination sharing person specific data situations aggregate statistical information rely transfer detailed transaction information time public information available data linked provide electronic person organization 
#@Pierangela Samarati, Latanya Sweeney
#t2009
#c9
#index10
#id10
#%
#!

#*mining association rules sets items large databases given large database customer transaction consists items customer present cient algorithm association rules items algorithm incorporates management novel estimation pruning present results applying algorithm sales data obtained large shows ectiveness 
#@Rakesh Agrawal, Tomasz Imielinski, Arun N. Swami
#t2009
#c10
#index11
#id11
#%
#!

#*fast algorithms mining association rules consider problem discovering association items large database sales transactions present new algorithms solving known algo empirical evaluation shows known algorithms factors ranging small problems order large problems proposed algorithms hybrid algorithm called scale scales number transactions scale properties respect number items database 
#@Rakesh Agrawal, Ramakrishnan Srikant
#t2009
#c11
#index12
#id12
#%
#!

#*privacy preserving data mining direction future data mining research development techniques incorporate privacy specifically address following primary task data mining development models data develop accurate models access precise information individual data records consider case building decision tree classifier training data values individual records 
#@Rakesh Agrawal, Ramakrishnan Srikant
#t2009
#c10
#index13
#id13
#%
#!

#*meta learning approach scalable accurate inductive learning research inductive learning problems relatively small amounts data age network computing likely orders magnitude data databases available various learning problems real world importance learning algorithms assume entire data set main memory feasible massive amounts data especially applications data mining approach large data set partition data set subsets run learning algorithm subsets combine results data distributed multiple sites network merging data location expensive propose investigate evaluate meta learning approach integrating results multiple learning processes approach utilizes machine learning guide integration identified main meta learning strategies strategies independent learning algorithms used generating classifiers strategy attempts reveal relationships learned classifiers prediction patterns strategy determine correct prediction classifiers different opinions various schemes strategies developed empirical results schemes obtain accurate classifiers classifiers trained data subsets implemented analyzed schemes parallel distributed environment demonstrate scalability 
#@Philip Kin-Wah Chan, Philip Kin-Wah Chan
#t2009
#c12
#index14
#id14
#%
#!

#*accuracy meta learning scalable data mining paper general approach scaling data mining applications come meta learning meta learning refers general strategy seeks learn combine number separate learning processes intelligent fashion meta learning architecture key meta learning strategy produce accurate final classification means meta learning architecture produce final accurate conventional learning algorithm applied available data second fast relative individual sequential learning algorithm applied massive databases examples operate reasonable time paper primarily issues related accuracy meta learning general strategy number empirical results presented meta learning feasible wide area network computing environments 
#@Philip K. Chan, Salvatore J. Stolfo
#t2009
#c13
#index15
#id15
#%
#!

#*distributed web mining using bayesian networks multiple data streams present approach bayesian net works distributed web log data streams approach learn local bayesian network site using local data site identifies observations likely evidence local non local variables set observations central site bayesian network central site using data local site local central bayesian networks combined obtain bayesian net work models entire data applied multiple data streams data difficult large response time scalability issues experimental results theoretical demonstrate feasibility approach presented 
#@R. Chen, Krishnamoorthy Sivakumar, Hillol Kargupta
#t2009
#c14
#index16
#id16
#%
#!

#*efficient mining association rules distributed databases sequential algorithms proposed mining association work mining association rules distributed database direct application sequential algorithms distributed databases effective requires large communication study efficient algorithm small number candidate sets requires messages support 
#@David Wai-Lok Cheung, Vincent Ng, Ada Wai-Chee Fu, Yongjian Fu
#t2009
#c15
#index17
#id17
#%
#!

#*secure multi party computation problems applications review open problems growth internet tremendous cooperative computation people computation tasks based private inputs computations occur parties example customers database queries contain private information organizations project satisfy organizations 
#@Wenliang Du, Mikhail J. Atallah
#t2009
#c16
#index18
#id18
#%
#!

#*secure multi party computational geometry general secure multi party computation problem parties private data respectively seek compute function inferred fi known theory secure multi party computation problem using approach 
#@Mikhail J. Atallah, Wenliang Du
#t2009
#c17
#index19
#id19
#%
#!

#*secure protocol computing dot products clustered distributed environments dot products form basis various applications ranging scientific computations commercial applications data mining transaction processing typical scientific computations sparse iterative use matrix vector products viewed dot products sparse vectors database applications dot products form operations widespread use clustered distributed operations increasingly performed hosts traditional data hosts entire computation example domain decomposition based sparse entire solution easily values net applications dot products performed hosts vectors need compute dot product cases need secure dot large computational requirements underlying applications highly desirable secure protocols minimal overhead original algorithm finally nature dot products limited amounts information parties detect entry party vector simply vector particular location given constraints traditional protocols generally significant computational communication paper present extremely efficient sufficiently secure protocol computing dot product vectors using linear techniques using experimental results demonstrate superior performance terms computational overhead numerical security overhead party dot product computation using high end workstations connected approaches multiple dot product average relative error dot products large number random normalized vectors roughly 
#@Ioannis Ioannidis, Ananth Grama, Mikhail Atallah
#t2009
#c18
#index20
#id20
#%
#!

#*privacy preserving distributed mining association rules partitioned data data mining extract important knowledge data collections collections various parties privacy concerns directly sharing data types data paper addresses secure mining partitioned data methods techniques minimize information shared adding little overhead mining task 
#@Murat Kantarcioglu, Chris Clifton
#t2009
#c15
#index21
#id21
#%
#!

#*distribution semantics new problem class distributed knowledge discovery fi research project use mobile data sources distributed knowledge current approaches suitable aim infer global model approximate results single data treat distribution technical issue distribution itself meaning models depend context main contribution paper identification practically relevant new problem class distributed knowledge discovery addresses semantics problem class proper framework important applications knowledge discovery process results process outline novel solution called knowledge discovery models uses models primary input combines content driven context driven finally discuss challenging research questions new problem 
#@Rdiger Wirth, Michael Borth, Daimlerchrysler Ag
#t2009
#c20
#index22
#id23
#%
#!

#*data satisfy privacy constraints data individuals entities collected widely data contain information explicitly identifies individual social security number data contain kinds personal information date code potentially identifying linked available data sets data shared business reasons paper addresses important issue preserving anonymity individuals entities data dissemination process explore preserving anonymity use potentially identifying portions data extend earlier works area various dimensions satisfying privacy constraints considered usage data allows optimize process preserving privacy specified usage particular investigate privacy transformation context data mining applications like building classification regression models second work improves previous approaches allowing data combined exploration solution space using genetic algorithm framework allow data useful intended purpose satisfying privacy constraints general terms privacy data transformation generalization predictive modeling 
#@Vijay S. Iyengar
#t2009
#c3
#index23
#id24
#%13,24,25,26,27,28,29,30,31,32,33,34,35,36
#!

#*comparing methods microdata basis information loss disclosure risk present paper empirical comparison microdata continuous categorical microdata identification experiments optimize disclosure relevant methods continuous generic information loss measures targeted data uses defined continuous categorical 
#@Josep Domingo-ferrer, Josep M. Mateo-sanz, Vicen? Torra
#t2009
#c21
#index24
#id25
#%
#!

#*supervised unsupervised discretization continuous features supervised machine learning algo require discrete feature space paper review previous work feature discretization identify ing characteristics methods empirical evaluation compare method entropy based based methods naive bayes algorithm features ing entropy based method fact tested datasets naive bayes cases performance induction algo improved features advance experiments performance interesting phenomenon considering ing features 
#@James Dougherty, Ron Kohavi, Mehran Sahami
#t2009
#c22
#index25
#id26
#%
#!

#*genetic algorithms search optimization machine learning book informal tutorial fashion techniques mathematical tools research results enable students apply genetic algorithms problems fields major concepts illustrated running examples major algorithms illustrated programs prior knowledge assumed minimum programming mathematics background required 
#@David E. Goldberg, John H. Holland
#t2009
#c23
#index26
#id27
#%
#!

#*use contextual information feature ranking discretization classification rules decision trees examples important problem features weak features process highly desirable numeric features need rule generation present new approach problems traditional techniques make use feature based information theoretic statistical correlation feature class instead assign features finding feature class discrimination context features used rank features select feature subset numeric variables experience benchmark example sets demonstrates new approach powerful alternative traditional methods paper new technical issues arise approach 
#@Se June Hong
#t2009
#c24
#index27
#id28
#%
#!

#*microdata files collect types data restrictions use microdata limited agents working secure systems restrictions affect public policy access summary statistics creation microdata requirements sufficient utility paper describes general methodology producing public use data files allows uses methodology quantitative data using noise approach necessary employs methodology major advantages scheme allows obtaining precise estimates possible known schemes addition distortion applied subset estimates file nearly identical file paper provides theoretical methodology results actual application using examples 
#@Jay J. Kim, William E. Winkler, Bureau Of The Census
#t2009
#c25
#index28
#id29
#%
#!

#*measures disclosure risk harm disclosure difficult topic definition disclosure depends context anonymity sensitive information disclosure occur information paper disclosure issues linking record learning sensitive information linking released record linked determines disclosure risk information linked released record determines disclosure harm harm record identified sensitive value inferred paper measures disclosure risk harm learned studied implications data release policies given 
#@Diane Lambert
#t2009
#c26
#index29
#id30
#%
#!

#*microdata release today great dissemination sharing information past released information statistical form situations today release specific data microdata order anonymity entities called information refers data explicit names addresses numbers identifying data provides anonymity released information contains data date code linked available information identify information intended disclosure paper address problem microdata anonymity data refer approach based definition anonymity table provides anonymity attempts link explicitly identifying information content map information entities illustrate anonymity provided information released using generalization techniques introduce concept minimal generalization captures property release process data needed achieve anonymity present algorithm computation generalization discuss possible preference policies choose different minimal 
#@Pierangela Samarati
#t2009
#c15
#index30
#id31
#%
#!

#*privacy information anonymity generalization today great dissemination sharing person situations aggregate statistical information rely detailed transaction time public information data provide electronic person organization 
#@Pierangela Samarati, Latanya Sweeney
#t2009
#c27
#index31
#id32
#%
#!

#*identification disclosure prediction disclosure microdata statistical disclosure identification disclosure prediction disclosure compared identification disclosure prediction disclosure sampling takes place cases prediction disclosure identification disclosure does small probability present disclosure problems different release aggregate statistics finally estimation population using poisson model considered 
#@C.J. Skinner
#t2009
#c28
#index32
#id33
#%
#!

#*algorithm rank based allocation best paper reports work past years using rank based allocation new evidence arguments presented suggest according rank superior fitness ranking used slow search speed increase search speed appropriate furthermore use ranking provides degree control possible fitness 
#@Darrell Whitley
#t2009
#c29
#index33
#id34
#%
#!

#*statistical disclosure control practice aim book discuss various aspects associated personal business data collected sources problem present data form useful statistical research provide sufficient protection individuals data refer major book concerned define disclosure problem deal practical 
#@L Willenborg, T De Waal
#t2009
#c30
#index34
#id35
#%
#!

#*elements statistical disclosure control statistical disclosure control producing statistical data released external researchers book methodology area microdata individual data data book attempts develop theory called paradigm statistical data way data minimum information loss book discusses data information loss measured data near optimal way measure safety information loss data data matter solving optimization problem problems discussed book turn hard problems solved approximately authors present new results published book description area closed fully explored book book useful social involved personal business data statistical use operations researchers interested optimization problems involved particularly challenges present statistical methods statistics main field research statistical disclosure control project project 
#@L Willenborg, T De Waal
#t2009
#c30
#index35
#id36
#%
#!

#*disclosure risk microdata protection paper describes methods data include rank noise describes enhanced methods identification using probabilistic record linkage empirical comparisons use variants framework measuring information loss identification risk introduced 
#@William E. Yancey, William E. Winkler, Robert H. Creecy
#t2009
#c31
#index36
#id37
#%
#!

#*document approach static index pruning text retrieval systems present static index pruning method used ad hoc document retrieval tasks follows approach decide posting given term index decision based term contribution document text collection global language model technique used decrease size index decrease retrieval ectiveness allows make index small main memory single large text collections containing millions documents results great efficiency superior earlier pruning methods average response time document collection 
#@Stefan Buttcher, CLA Clarke
#t2009
#c32
#index37
#id38
#%38,39,40,41,42,43,44,45,46,47,48,49,50
#!

#*query evaluation using exhaustive evaluation ranked queries expensive particularly small subset overall ranking required queries contain common terms gives rise techniques dynamic query pruning methods parts usual exhaustive evaluation generating good set answers query work propose new pruning methods make use impact sorted indexes compared exhaustive evaluation new methods reduce computation performed reduce memory required reduce data disk time allow performance guarantees terms precision mean average precision strong experiments using trec terabyte collection queries 
#@VO NGOC ANH, Alistair Moffat
#t2009
#c33
#index38
#id39
#%
#!

#*efficient computation multiple language model multiple language model generally considered computationally expensive practical purposes efficient multinomial approach model properties little known retrieval effectiveness model high cost execution paper efficient implementation model achieved resulting method comparable terms efficiency standard term matching algorithms vector space model multinomial language model 
#@Leif Azzopardi, David E. Losada
#t2009
#c34
#index39
#id40
#%
#!

#*efficient phrase querying index search engines need evaluate queries extremely fast challenging task given quantities data indexed significant queries search engines involve phrases paper consider phrase queries efficiently supported low disk previous research shown phrase queries rapidly evaluated using nextword indexes indexes large conventional inverted files propose combination nextword indexes inverted files solution problem experiments combined use nextword index conventional inverted file allow evaluation phrase queries time required evaluate queries inverted file space overhead size inverted file time available increases disk requirements general terms indexing query evaluation keywords inverted indexes nextword indexes evaluation efficiency index size phrase query 
#@Dirk Bahle, Hugh E. Williams, Justin Zobel
#t2009
#c35
#index40
#id41
#%51,52,53,54,55,56,57,45,58,46,59,60,61,62
#!

#*techniques efficient query expansion fi query expansion known method improving ectiveness information ective query expansion methods rely costly retrieval processing documentsfi explore alternative methods reducing query evaluation costs propose new method based summary document method allows query times faster previously ectiveness standard 
#@Bodo Billerbeck, Justin Zobel
#t2009
#c36
#index41
#id42
#%
#!

#*static index pruning information retrieval systems introduce static index pruning methods significantly reduce index size information retrieval systems investigate uniform term based methods selected entries index effect retrieval results uniform pruning fixed threshold index entries contribution relevance scores bounded given threshold index term based pruning threshold determined term vary term term experimental evidence level compression term based pruning outperforms uniform pruning various measures precision present theoretical experimental evidence term based pruning scheme possible index greatly retrieval results good based index 
#@Aya Soffer, David Carmel, Doron Cohen, Ronald Fagin, Eitan Farchi, Michael Herscovici, Yoelle S. Maarek
#t2009
#c35
#index42
#id43
#%63,64,65,66,67,68,46,69
#!

#*information theoretic approach automatic query expansion techniques automatic query expansion retrieved documents shown improving retrieval effectiveness large collections rely empirical cross comparisons using ideas information theory present computationally simple method assigning scores candidate expansion terms scores used select weight expansion terms rocchio framework query compare ranking information theoretic query expansion versus ranking query expansion techniques showing achieves better retrieval effectiveness performance measures discuss effect retrieval effectiveness main parameters involved automatic query expansion data query difficulty number selected documents number selected terms interesting relationships 
#@Claudio Carpineto, Renato de Mori, Giovanni Romano, Brigitte Bigi
#t2009
#c37
#index43
#id44
#%
#!

#*improving web search efficiency based static pruning method fast continuous growth volume indexed documents web great search engines true search ectiveness time space paper present index pruning technique targeted search engines addresses issue adopt new pruning strategy greatly reducing size search engine indices experiments using real search engine technique reduce indices age costs traditional methods loss retrieval precision minimum compared indices size compression compression rate higher original size results indicate reduction storage overhead query processing time reduced nearly original time loss age precision new method yields improve compared best known static ing method search engine indices addition technique orthogonal underlying search algorithms adopted search engine 
#@Edleno Silva de Moura, Celia Francisca dos Santos, Daniel R. Fernandes, Altigran Soares da Silva, Pavel Calado, Mario A. Nascimento
#t2009
#c38
#index44
#id45
#%
#!

#*self indexing inverted files fast text retrieval query processing costs large text databases need retrieve scan inverted list query term query response time boolean queries informal ranked queries dramatically reduced little cost terms storage internal index inverted list method applied retrieval collection nearly million short documents experimental results strategy size inverted file boolean queries terms reduce processing time previous cost ranked queries terms evaluated little previous time little loss retrieval ectiveness 
#@Alistair Moffat, Justin Zobel
#t2009
#c37
#index45
#id46
#%70,71,63,72,73,74,75,76,77,78,79,80,81,82,83,84
#!

#*document retrieval frequency sorted indexes ranking techniques ective finding answers document collections expensive evaluate propose evaluation technique uses early recognition documents likely highly ranked reduce costs test data queries evaluated memory standard implementation retrieval ectiveness cpu time disk traffic dramatically reduced inverted indexes explicitly support technique principle index design inverted lists sorted document frequency document number method experimentally reduces cpu time disk traffic original requirement frequency sorting lead net reduction index size index compressed 
#@Michael Persin, Justin Zobel, Ron Sacks-Davis
#t2009
#c39
#index46
#id47
#%85,86,66,87,45,81,88,83,84
#!

#*university trec experiments web robust terabyte tracks trec test scalable information retrieval framework tracks mixed query task web track employ decision mechanism selecting appropriate retrieval approaches query basis robust track order queries use pre retrieval performance predictors weighting function recommender mechanism test new training approach automatic term frequency parameters terabyte track employ distributed version test effectiveness techniques using text pseudo query expansion selecting different weighting models query 
#@Vassilis Plachouras, Ben He, Iadh Ounis
#t2009
#c40
#index47
#id48
#%
#!

#*trec article presents text retrieval research results areas automatic ad hoc retrieval information filtering large collections interactive retrieval automatic ad hoc retrieval runs submitted medium title description short title run combination long run title description medium short runs average precision run higher submitted run recently high score topic short run led discovery indexing procedures subject field times documents indexed use field explicitly ad hoc track runs index results presented average reduced area adaptive filtering focused problems starting assumed history relevance judgments topic having define threshold retrieval runs database submitted collections area interactive retrieval comparisons relevance feedback 
#@S.E. Robertson, S. Walker, M. Beaulieu, Peter Willett
#t2009
#c41
#index48
#id49
#%
#!

#*compression inverted indexes fast query evaluation compression reduces size indexes time needed evaluate queries paper compression inverted lists document postings store position frequency indexed terms considering approaches improving retrieval efficiency better implementation better choice integer compression schemes propose simple known integer compression schemes experimentally lead significant time second explore impact choice compression scheme retrieval efficiency experiments large collections data results use simple byte aligned codes query evaluation time compared compact compression schemes index memory byte aligned codes result faster query evaluation does index cost data memory cpu cache compressed index index byte aligned schemes space overhead compact schemes result indexes size collection byte aligned scheme conclude fast byte aligned codes used store integers inverted lists 
#@Falk Scholer, Hugh E. Williams, John Yiannis, Justin Zobel
#t2009
#c35
#index49
#id50
#%45,46,59,61
#!

#*query evaluation strategies optimization discusses query evaluation strategies used large text retrieval systems term time document time describes optimization techniques reduce query evaluation costs presents simulation results compare performance optimization techniques applied natural language query evaluation 
#@Howard Turtle, James Flood
#t2009
#c42
#index50
#id51
#%
#!

#*vector space ranking ective early termination considerable research effort improving effectiveness information retrieval systems techniques relevance feedback expansion provide better quality queries tested standard evaluation enhancements cost evaluating queries paper consider issue improve cost effectiveness searching new inverted file structure using weights provides superior retrieval effectiveness compared conventional inverted file structures early termination heuristics employed able similar effectiveness levels computational cost provide better cost performance previous inverted file 
#@Vo Ngoc Anh, Owen de Kretser, Alistair Moffat
#t2009
#c43
#index51
#id52
#%
#!

#*techniques nextword indexes queries text search engines ranked boolean phrase querying powerful technique searches expensive implement conventional indexes work nextword index proposed structure designed phrase queries nextword indexes relatively large paper introduce new techniques nextword indexes contrast index compression schemes techniques allow resolution phrase queries false match experimentally novel techniques lead index size 
#@Dirk Bahle, Hugh E. Williams, Justin Zobel
#t2009
#c36
#index52
#id53
#%
#!

#*phrase querying browsing text databases search systems querying large document example web search based information retrieval principles systems eficient effective finding answers user information needs expressed informal ranked structured boolean queries phrase querying browsing additional techniques augment conventional querying tools paper propose phrase nextword index eficient structure phrase based searching consideration search terms evaluated query plan order evaluation plan reduce query evaluation costs factor conclude phrase querying browsing nextword indexes ordered query plan used browsing querying phrase querying practical large text collections 
#@Dirk Bahle, Hugh E. Williams, Justin Zobel
#t2009
#c44
#index53
#id54
#%
#!

#*interactive internet search keyword query mechanisms compared article compares search effectiveness based internet search google based search yahoo search browser means focus search load using task effort user various search states independent 
#@Peter Bruza, Robert McArthur, Simon Dennis
#t2009
#c35
#index54
#id55
#%
#!

#*relevance ranking term queries search systems querying large document example web search based information retrieval principles systems eficient effective finding answers user information needs expressed informal ranked structured boolean queries phrase querying browsing additional techniques augment conventional querying tools paper propose phrase nextword index eficient structure phrase based searching consideration search terms evaluated query plan order evaluation plan reduce query evaluation costs factor conclude phrase querying browsing nextword indexes ordered query plan used browsing querying phrase querying practical large text collections 
#@CLA Clarke, Gordon V. Cormack, EA Tudhope
#t2009
#c45
#index55
#id56
#%
#!

#*phrase recognition expansion short precision biased queries based query log paper examine question query parsing world wide web queries present novel method phrase recognition given training corpus approximately million web queries context free grammar em algorithm used estimate parameters probabilistic context free grammar developed fi use compute user query linguistic structure word 
#@Erika F. de Lima, Jan O. Pedersen
#t2009
#c35
#index56
#id57
#%
#!

#*results challenges web search evaluation million page web created enable meaningful evaluation web search systems techniques collection used evaluation framework text retrieval trec provide answers questions information result better rankings longer queries result better answers trec systems work web data associated described extended preliminary results comparison trec systems working known web search systems working current web standard document rankings produced public web search engines means state art 
#@David Hawking, Nick Craswell, Paul B. Thistlewaite, Donna Harman
#t2009
#c46
#index57
#id58
#%
#!

#*scalable browsing large collections case study phrase browsing techniques use phrases large information collection browsing paper describes uses automatically constructed facilitate browsing large phrases extracted text using syntactic processing grammar induction interface robust easy 
#@Gordon W. Paynter, Ian H. Witten, Sally Jo Cunningham, George Buchanan
#t2009
#c47
#index58
#id59
#%
#!

#*searching web public queries studying actual web searching public large analyzed million web queries users search engine people use search terms modified queries view web pages use advanced search small number search terms used high frequency great terms unique language web queries queries en rank compared data large studies web study provides insight public choices web 
#@A Spink, D Wolfram, MBJ Jansen, T Saracevic
#t2009
#c48
#index59
#id60
#%
#!

#*index structures efficient phrase querying text retrieval systems used documents large text collections using queries consisting words word sequences 
#@Hugh E. Williams, Justin Zobel, Phil Anderson
#t2009
#c49
#index60
#id61
#%
#!

#*managing indexing documents images fully updated second edition highly managing authors provide coverage state art techniques indexing data field work large quantities information book essential authoritative theoretical resource practical guide storage access challenges covers compression indexing application web digital libraries details powerful techniques supported authors storing retrieving text images textual images source code available web 
#@Ian H. Witten, Alistair Moffat, Timothy C. Bell
#t2009
#c50
#index61
#id62
#%
#!

#*exploring similarity space ranked queries used relevant documents text database ranked query list terms specified documents closely match query order similarity crucial ranked querying use similarity heuristic mechanism assigns numeric score closely document query explore range similarity heuristics described 
#@Justin Zobel, Alistair Moffat
#t2009
#c51
#index62
#id63
#%
#!

#*optimization inverted vector searches simple algorithm presented increasing efficiency information retrieval searches implemented using inverted files optimization algorithm employs knowledge methods used weighting document query terms order examine inverted lists possible extension basic algorithm allows greatly increased performance optimization cost retrieval effectiveness experimental runs examining different term weighting models showing optimization possible 
#@Chris Buckley, A. F. Lewit
#t2009
#c35
#index63
#id64
#%
#!

#*new retrieval approaches using smart trec smart information retrieval project completely automatic approaches retrieval large quantities work trec performing runs ad hoc text interactive language years smart project university interested analysis search heterogeneous text databases vocabulary vary widely 
#@Chris Buckley, Amit Singhal, Mandar Mitra
#t2009
#c52
#index64
#id65
#%
#!

#*indexing latent semantic analysis new method automatic indexing retrieval described approach advantage implicit higher order structure association terms documents semantic structure order improve detection relevant documents basis terms queries particular technique used singular value decomposition large term document matrix set factors original matrix linear combination documents represented item vectors factor weights queries represented pseudo document vectors formed weighted terms documents threshold values initial tests completely automatic method retrieval promising 
#@Scott C. Deerwester, Susan Dumais, Thomas K. Landauer, George W. Furnas, Richard A. Harshman
#t2009
#c39
#index65
#id66
#%
#!

#*retrieving records text using statistical ranking statistically based ranked retrieval records using keywords provides advantages traditional boolean retrieval methods especially end users approach retrieval seen widespread use large retrieval systems feasibility retrieval methodology research produce fast search techniques using ranking algorithms test results large databases end users results response times order text user users able consistently obtain good search results training additional work new indexing techniques create inverted files large databases using techniques use sorting require working space size input text produce indices input text size 
#@Donna Harman, C Gerald
#t2009
#c53
#index66
#id67
#%
#!

#*text indexing based lexical relations application software libraries contrast kinds libraries software libraries need conceptually component main users desired component implementation details software reuse enhanced conceptually large libraries software components paper present tool allows building large software libraries software components focus indexing component attributes natural language indexing method based words occurrences uses extract occurrence knowledge extracting potential attributes textual documents conceptually relevant selected according power scales noise context words fully automated indexing tool keyword based tools understanding document knowledge based tools indexing component fully implemented results given paper 
#@Y. S. Maarek, F.Z.Smadja
#t2009
#c54
#index67
#id68
#%
#!

#*fast ranking limited space ranking techniques long suggested alternatives methods searching document computing ranking cost performing search terms memory space processing consider resources required method ranking application indexing selection space time required ranking substantially 
#@Alistair Moffat, Justin Zobel
#t2009
#c55
#index68
#id69
#%
#!

#*overview text retrieval conference trec paper serves introduction research described main task ad hoc retrieval defined details test collections evaluation methodology used trec provides overview ad hoc retrieval addition main ad hoc task trec tracks tasks focus research particular 
#@EM Voorhees, Donna Harman
#t2009
#c56
#index69
#id70
#%
#!

#*data compression retrieval systems describes compression methods components text systems text databases cd rom topics discussed include storage media structures text retrieval including indexes inverted files compression tools memory requirements retrieval ranking information retrieval 
#@Timothy C. Bell, Alistair Moffat, Craig G. Nevill-Manning, Ian H. Witten, Justin Zobel
#t2009
#c57
#index70
#id71
#%
#!

#*model based concordance compression authors discuss concordance compression using framework compression theory begin creating mathematical model concordance generation use optimal compression engines coding actual compression context static information retrieval compression tasks compression building needed processing query directly response time use extensive costly compression provided fast methods possible compression applied files text concordance needed possibly short pieces random means exact locations use adaptive methods based tables systematically change end file speed encoding concordance compression conceptually modern approach data compression testing effectiveness models 
#@Abraham Bookstein, Shmuel T. Klein, Timo Raita
#t2009
#c58
#index71
#id72
#%
#!

#*database systems clear theory design broad coverage models real systems date introduction modern database technologies result leading introduction database systems new problems new manual students practice design implementation real world examples serve practical database concepts edition maintains coverage popular database topics including sql security data mining contains new chapter web programming databases 
#@Ramez Elmasri, Shamkant B. Navathe
#t2009
#c50
#index72
#id73
#%
#!

#*access methods text paper compares text retrieval methods intended systems requirements environment discussed retrieval methods database systems information retrieval systems examined classify methods examine interesting class attempts speed retrieval special purpose hardware presented issues approximate matching compression discussed comparison examined methods presented file method discussed 
#@Christos Faloutsos
#t2009
#c59
#index73
#id74
#%
#!

#*files access method documents performance evaluation file access method text retrieval studied according method documents stored sequentially text file documents stored file serves filter retrieval helps large number documents paper methods creating studied based word coding closed form derived false probability methods factors affect studied performance comparisons methods based provided 
#@Christos Faloutsos, Stavros Christodoulakis
#t2009
#c37
#index74
#id75
#%
#!

#*optimal source codes distributed integer theta theta probability assignment set integers arbitrary real number theta optimal binary source code probability assignment constructed follows integer theta theta theta represent integer code followed single code using log log log optimal code integers codes 
#@R Gallager, D Van Voorhis
#t2009
#c60
#index75
#id76
#%
#!

#*storing text retrieval systems cd rom compression considerations cd rom storage medium text databases question maximum size database contained medium example problem storing cd rom examined paper text database long cd rom addition dictionary concordance needed access data stored constraint material desirable material difficult software provided approaches compression various files compression text related problem data specifically shown simple models text generation encoding produces representation 
#@Shmuel T. Klein, Abraham Bookstein, Scott C. Deerwester
#t2009
#c35
#index76
#id78
#%
#!

#*compression indexes information large text databases paper describes combination compression methods used reduce size inverted indexes large text databases methods prefix run length encoding novel family numeric representations called coding using compression methods different text sources version sample journal stories compressed index size original text numbers index time required fully time needed postings research effort handle large text databases parallel 
#@Gordon Linoff, Craig Stanfill
#t2009
#c35
#index77
#id79
#%
#!

#*development stemming algorithm stemming algorithm procedure reduce words common form useful areas computational linguistics information retrieval work form algorithm applications certain linguistic problems common stemming procedure basis evaluation previous attempts deal problems paper discusses theoretical practical attributes stemming algorithms new version context sensitive match stemming algorithm english developed use library information transfer general applications proposed major linguistic problem stemming variation spelling discussed feasible solutions sample results methods 
#@Julie B. Lovins
#t2009
#c61
#index78
#id80
#%
#!

#*inverted index implementation working implementation inverted index suitable real time applications described based hash addressed random access organization variable length records index structure processing algorithms developed simulation model typical simulation results presented critical parameters identified solution inverted key identification problem proposed time software directly dependent main file record 
#@Ken J. McDonell
#t2009
#c62
#index79
#id81
#%
#!

#*compression sparse text retrieval systems typically use inverted identify documents contain words documents containing words quickly word occurrences large usually sparse variety compression consider techniques encoding different code used 
#@Alistair Moffat, Justin Zobel
#t2009
#c35
#index80
#id82
#%
#!

#*memory efficient ranking fast effective ranking collection documents respect query requires structures including vocabulary inverted file entries term weights document lengths array partial similarity address tables inverted file entries documentsfi structures array document lengths array components frequently ranked query crucial 
#@Alistair Moffat, Justin Zobel, Ron Sacks-Davis
#t2009
#c63
#index81
#id83
#%
#!

#*multi key access methods based coding techniques single level level indexed schemes retrieval presented compared descriptors formed using coding techniques stored using technique fast batch algorithm cost inverted file disk access record presented large data files shown level implementation generally efficient queries small number matching records queries specify values potential problem level implementation costs blocks records match query individual records blocks approach problem set descriptors based pairs indexed terms approach presented analyzed 
#@R. Sacks-Davis, A. Kent, K. Ramamohanarao
#t2009
#c64
#index82
#id84
#%
#!

#*introduction modern information retrieval new technology allows design information retrieval analyze process store retrieve specific resources matching particular user needs clear practical text theory techniques tools critical making information retrieval work completely second edition incorporates rapidly field including multimedia information retrieval user interfaces digital libraries coverage comprehensive including classification subject indexing ing vocabulary control cd rom online information retrieval multimedia hypertext expert systems natural language processing user interface systems internet world wide web digital library environments illustrated examples ideal students library information studies advance knowledge future information 
#@Gerard Salton, Michael McGill
#t2009
#c50
#index83
#id85
#%
#!

#*efficient indexing technique text database systems text database systems require index allow fast access documents based content propose inverted file indexing scheme based compression scheme allows users retrieve documents using words occurring documents sequences words statistical ranking techniques compression methods chosen ensure storage requirements small dynamic update assumption make sufficient main memory available support memory vocabulary given assumption method requires access query term identify answers queries 
#@Justin Zobel, Alistair Moffat, Ron Sacks-Davis
#t2009
#c65
#index84
#id86
#%
#!

#*data compression text retrieval systems describes compression methods components text systems text databases cd rom topics discussed include storage media structures text retrieval including indexes inverted files compression tools memory requirements retrieval ranking information retrieval 
#@Timothy C. Bell, Alistair Moffat, Craig G. Nevill-Manning, Ian H. Witten, Justin Zobel
#t2009
#c39
#index85
#id87
#%
#!

#*inverted vector searches review use inverted files best match searching information retrieval systems 
#@Chris Buckley, Alan F. Lewit
#t2009
#c43
#index86
#id88
#%
#!

#*coding compression text retrieval systems described compression models use text retrieval systems authors discuss coding methods use models results scheme identical compression times faster main features implementation complete coding high speed implementation particularly suited slow devices cd rom answering query requires disk access term query disk access answer words numbers indexed words built compressed databases 
#@Alistair Moffat, Justin Zobel
#t2009
#c58
#index87
#id89
#%
#!

#*review use inverted files best match searching information retrieval systems use inverted files calculation similarity coefficients types matching function discussed context document retrieval systems critical evaluation presented range algorithms described matching documents queries particular attention computational efficiency various procedures improved search heuristics given cases suggested algorithms implemented sufficiently efficiently nearest searching standard retrieval 
#@Shirley A. Perry, Peter Willett
#t2009
#c66
#index88
#id90
#%
#!

#*fast high quality multilevel scheme partitioning graphs recently number researchers investigated class graph partitioning algorithms reduce size graph vertices edges partition smaller graph construct partition original graph early work clear multilevel techniques held great known consistently produce high quality partitions graphs arising wide range application domains investigate ectiveness di erent choices partition graph refinement particular present new heuristic called heuristic size partition graph small factor size final partition obtained multilevel refinement present faster variation lin algorithm test scheme large number graphs arising various domains including finite element methods linear programming experiments scheme produces partitions consistently better produced spectral partitioning schemes substantially smaller time scheme used compute reducing orderings sparse matrices produces orderings substantially smaller widely used multiple minimum degree algorithm 
#@G Karypis, Vipin Kumar
#t2009
#c67
#index89
#id91
#%90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119
#!

#*fast multilevel implementation recursive spectral bisection partitioning unstructured problems unstructured meshes used large scale scientific engineering problems including finite volume methods computational dynamics finite element methods structural analysis unstructured problems solved distributed memory parallel computers data structures partitioned distributed processors solved efficiently partitioning maximize load balance minimize 
#@Stephen T. Barnard, Horst D. Simon
#t2009
#c68
#index90
#id92
#%
#!

#*algorithm partitioning nodes graph let graph having nodes edges consider problem partitioning subsets given sizes respectively way number edges different subsets minimal obtain heuristic solution solution linear programming problem 
#@Earl R. Barnes
#t2009
#c69
#index91
#id93
#%
#!

#*graph bisection algorithms good average case behavior paper polynomial time algorithm input graph outputs minimum bisection graph output algorithm course high probability natural classes graphs particular sufficiently algorithm finds minimum bisection simple graphs nodes bisection example algorithm graphs nodes bisection algorithm graph bisection heuristics heuristics np complete problems 
#@Thang Nguyen Bui, Soma Chaudhuri, Frank Thomson Leighton, Michael Sipser
#t2009
#c70
#index92
#id94
#%
#!

#*geometric spectral partitioning investigate new method partitioning graph equal pieces combine ideas recently suggested partitioning algorithms spectral bisection uses matrix associated graph geometric bisection applies graphs space fi new method does require geometric produces partitions better spectral geometric 
#@Tony F. Chan, John R. Gilbert, Shang-Hua Teng
#t2009
#c50
#index93
#id95
#%
#!

#*parallel algorithms dynamically partitioning unstructured grid partitioning method choice wide variety naturally parallel problems computational grid grid itself changes simulation ability parallel achieving higher algorithms suitable parallel dynamic load balancing partition unstructured computational load 
#@Pedro C. Diniz, Steve Plimpton, Bruce Hendrickson, Robert W. Leland
#t2009
#c68
#index94
#id96
#%
#!

#*linear time heuristic improving network partitions cut heuristic provides efficient solution separating network vertices separate partitions effort number contain nodes partition heuristic specifically handle large complex networks contain multi weighted worst case computation time heuristic overall size network practice seen number typically required converge value typically small key factor obtaining linear fact algorithm node time attempt reduce current cut set maximum possible value times maximum cut set reduction negative algorithm allows local currently movement node appropriate updated represent use simple data structures allows ability avoid searching best updates need updated data structures adding efficient nature algorithm heuristic contains balance allows user control size created partitions 
#@C. M. Fiduccia, R. M. Mattheyses
#t2009
#c71
#index95
#id97
#%
#!

#*finding clusters circuit partitioning fundamental role hierarchical layout systems identifying strongly connected clusters logic significantly reduce circuit total interconnection length finding cluster partition np complete authors propose fast heuristic algorithm based simple local criterion able prove highly structured clusters algorithm correspond high probability natural clusters application large scale real world shows method number cut reduced compared standard approach 
#@Jorn Garbers, Hans Jurgen Promel, Angelika Steger
#t2009
#c72
#index96
#id98
#%
#!

#*evolution minimum degree ordering algorithm past years implementation minimum degree algorithm study important enhancements article various enhancements historical development provide experiments showing effective improving execution time algorithm present widely used implementations algorithm quality ordering provided implementations sensitive initial ordering example changing input ordering lead increase decrease factor cost subsequent numerical factorization effective strategy experiments illustrate importance developing strategy 
#@Alan George, Joseph W.H. Liu 1989
#t2009
#c50
#index97
#id99
#%
#!

#*geometric mesh partitioning implementation experiments investigate method mesh equal pieces method novel feature geometric mesh based theoretical work showed certain classes finite element meshes good geometric method quite simple implement code method quite 
#@John R. Gilbert, Gary L. Miller, Shang-Hua Teng
#t2009
#c73
#index98
#id100
#%
#!

#*heuristic algorithms automatic graph partitioning practical implementations finite element method distributed memory multi systems use partitioning tools mesh sub meshes roughly equal size graph partitioning algorithms implementing distributed sparse matrix methods domain decomposition techniques structured problems parallel computers propose class algorithms based level set number center nodes 
#@T. Goehring, Y. Saad
#t2009
#c74
#index99
#id101
#%
#!

#*highly scalable parallel algorithms sparse matrix factorization paper scalable parallel algorithms sparse matrix factorization analyze performance scalability present experimental results processors parallel analysis experimental results demonstrate algorithms substantially improve state art parallel direct solution sparse linear systems terms scalability overall performance known fact dense matrix factorization scales implemented efficiently parallel computers paper present algorithms factor wide class sparse matrices including arising dimensional finite element problems scalable dense matrix factorization algorithms variety parallel architectures algorithms communication overhead scalable previously known parallel formulation sparse matrix factorization paper discuss cholesky factorization symmetric positive matrices algorithms solving sparse linear problems gaussian elimination matrices symmetric structure implementation sparse cholesky factorization algorithms medium size structural engineering linear programming problems best knowledge highest performance obtained sparse cholesky factorization 
#@Anshul Gupta, G Karypis, Vipin Kumar
#t2009
#c75
#index100
#id102
#%
#!

#*fast spectral methods ratio cut partitioning clustering ratio cut partitioning objective function successfully traditional cut partitioning style ratio cut heuristics achieved cost circuit partitioning hardware simulation applications authors theoretical optimal ratio cut partition cost second particular derived matrix present fast based methods computing heuristic ratio eigenvector second results better previous methods average primary efficient clustering method based second eigenvector successful difficult input classes design literature directions future work considered 
#@Lars W. Hagen, Andrew B. Kahng
#t2009
#c72
#index101
#id103
#%
#!

#*new approach ective circuit clustering complexity generation systems capabilities layout algorithms particularly partitioning module clustering needed problem size tractable existing optimization methods paper establish measure general metric evaluation clustering algorithms metric turn algorithm new self clustering method based random circuit efficiently captures good circuit clustering incorporated phase iterative partitioning strategy clustering method improves bisection average previous based methods 
#@Lars Hagen, Andrew B. Kahng
#t2009
#c76
#index102
#id104
#%
#!

#*parallel algorithms matrix computations shown mesh connected compute inverse matrix linear time algorithm based theorem known matrix steps involves unit distance message routing operations processor coding memory requirement processor independent shown algorithm systems linear equations linear time processors 
#@K. A. Gallivan, Michael T. Heath, Esmond Ng, James M. Ortega, Barry W. Peyton, R. J. Plemmons, Charles H. Romine, A. H. Sameh, Robert G. Voigt
#t2009
#c30
#index103
#id105
#%
#!

#*parallel nested dissection algorithm paper concerned parallel computation ordering symmetric positive sparse matrix purpose ordering limit enhance computation cholesky matrix use geometric approach nested dissection given em graph matrix space resulting algorithm implemented parallel memory 
#@Michael T. Heath, Padma Raghavan
#t2009
#c77
#index104
#id106
#%
#!

#*improved spectral graph partitioning algorithm mapping parallel computations efficient use distributed memory parallel requires computational load balanced processors way minimizes communication new domain mapping algorithm presented recent work ideas spectral graph theory applied problem generalization spectral graph bisection involves novel use multiple allow computation parts stage recursive decomposition resulting method suitable scientific computations like finite elements differences performed mesh architecture machines experimental results new method provides better previous spectral methods algorithm allows arbitrary weights vertices edges model computation communication new spectral lower bound graph bisection presented 
#@Bruce Hendrickson, Robert Leland
#t2009
#c67
#index105
#id107
#%
#!

#*multilevel algorithm partitioning graphs graph partitioning problem vertices graph sets specified sizes edges cross np complete problem arises important scientific engineering examples include mapping parallel computations ordering sparse matrix present multilevel algorithm graph partitioning graph sequence increasingly 
#@Bruce Hendrickson, Robert Leland
#t2009
#c78
#index106
#id108
#%
#!

#*parallel formulation interior point algorithms recent years interior point algorithms used successfully solving medium large size linear programming problems paper highly parallel formulation interior point algorithm key component interior point algorithm solution sparse linear equations using cholesky factorization performance parallel cholesky factorization determined communication overhead algorithm load processors parallel interior point algorithm use recently developed parallel algorithm communication overhead parallel algorithms cholesky factorization developed date computation depends shape elimination tree associated sparse factorization balance computation implemented evaluated di erent ordering algorithms algorithms lin spectral nested dissection yield balanced elimination trees greatly increase preliminary implementation achieves high processor size problems 
#@G Karypis, Anshul Gupta, Vipin Kumar
#t2009
#c79
#index107
#id109
#%
#!

#*analysis multilevel graph partitioning recently number researchers class algorithms based multilevel computational complexity provide excellent graph exist little theoretical analysis explain multilevel algorithms produce good present assumptions refinement phase 
#@G Karypis, Vipin Kumar
#t2009
#c79
#index108
#id110
#%
#!

#*parallel algorithm multilevel graph partitioning sparse matrix ordering paper present parallel formulation multilevel graph partitioning sparse matrix key feature parallel formulation proposed parallel algorithms partitions vertices graph parts overall graph mapping results substantially smaller communication 
#@G Karypis, Vipin Kumar
#t2009
#c80
#index109
#id111
#%
#!

#*separator theorem planar graphs let vertex planar graph prove partitioned sets vertex vertex vertices contains vertices exhibit algorithm finds partition time 
#@Richard J. Lipton, Robert E. Tarjan
#t2009
#c81
#index110
#id113
#%
#!

#*minimum degree algorithm multiple elimination widely used ordering scheme reduce operations sparse matrix computation minimum degree algorithm notion multiple elimination introduced conventional scheme discussed using grid model problem experimental results indicate modified version reducing property better original ordering algorithm requires time reduction ordering time problem dependent problems modified algorithm run times faster existing implementations minimum degree algorithm use external degree algorithm introduced 
#@Joseph W. H.
#t2009
#c82
#index111
#id114
#%
#!

#*automatic mesh partitioning paper describes efficient approach partitioning unstructured meshes occur naturally finite element finite difference methods approach makes use underlying geometric structure given mesh finds provably good partition random time applies meshes dimensions new method applications efficient sequential parallel algorithms large scale problems scientific computing overview paper written emphasis aspects approach detailed papers 
#@Gary L. Miller, Shang Teng, William Thurston, Stephen A. Vavasis
#t2009
#c83
#index112
#id115
#%
#!

#*unified geometric approach graph separators class graphs called overlap graphs proposed special cases overlap graphs include planar graphs nearest neighbor graphs earlier classes graphs associated finite element methods separator bound overlap graphs embedded dimensions result earlier separator results arguments based geometric properties separator bounds come randomized linear time randomized algorithms bounds best possible leading term 
#@Gary L. Miller, Shang-Hua Teng, Stephen A. Vavasis
#t2009
#c84
#index113
#id116
#%
#!

#*oriented approach partitioning large sparse graphs connectivity constraint paper consider problem partitioning large sparse graphs finite element meshes heuristic proposed allows partition connected balanced reasonable time minimize number goal build partitions graphs containing large numbers nodes edges practice algorithm iterative construction connected construction achieved exploring clusters nodes called use connectivity subsets low cost shown given complexity operations grows number edges examples given illustrate quality speed heuristic 
#@P Ciarlet, F Lamour
#t2009
#c85
#index114
#id117
#%
#!

#*combinatorial optimization comprehensive combinatorial optimization special emphasis theoretical results algorithms provably good performance contrast heuristics basis combinatorial optimization special topics level complete book contains material usually selects material suitable way book contains complete results appear book recent topics references provided book represents state art combinatorial optimization edition contains new chapter location problems area extremely active past years furthermore new material various topics new updates added reviews edition book combinatorial optimization example ideal operations second edition updates book documents relevant knowledge combinatorial optimization records problems algorithms define today read researchers students interested combinatorial optimization 
#@Vijay Chandru, M. R. Rao
#t2009
#c86
#index115
#id118
#%
#!

#*computing block form sparse matrix consider problem sparse matrix compute block form block form based decomposition bipartite graphs maximum matching discovered implementations algorithms compute block form provide computational results sparse matrices test collections applications block form 
#@Alex Pothen, Chin-Ju Fan
#t2009
#c87
#index116
#id119
#%
#!

#*partitioning sparse matrices graphs problem computing small vertex separator graph arises context computing good ordering parallel factorization sparse symmetric matrices approach computing vertex separators considered paper shown lower bounds separator sizes obtained terms matrix associated graph grid graphs computed products involving path graphs used compute good separators grid graphs heuristic algorithm designed compute vertex separator general graph computing separator graph eigenvector matrix using maximum matching compute vertex separator results quality separators computed spectral algorithm presented compared separators obtained automatic nested dissection lin algorithm finally report time required compute eigenvector consider accuracy eigenvector computed obtain good separators spectral algorithm advantage implemented medium size manner 
#@Alex Pothen, Horst D. Simon, Kan-Pu Liou
#t2009
#c77
#index117
#id120
#%
#!

#*spectral nested dissection fi spectral nested dissection algorithm computing orderings appropriate factorization sparse symmetric algorithm makes use spectral properties matrix associated given matrix compute evaluate quality spectral orderings respect measures elimination tree weight elimination trees tree spectral orderings compare quite 
#@Alex Pothen, Horst D. Simon, Lie Wang
#t2009
#c88
#index118
#id121
#%
#!

#*fast implementation spectral nested dissection authors novel spectral nested dissection algorithm novel algorithm computing orderings appropriate parallel factorization sparse symmetric matrices algorithm makes use spectral properties matrix associated given matrix compute separators authors evaluate quality spectral orderings respect measures elimination tree weight elimination trees tree use large structural analysis problems test cases demonstrate real applications spectral orderings compare quite commonly used orderings wide measures relatively long execution time 
#@Alex Pothen, Horst D. Simon, Lie Wang, Stephen T. Barnard
#t2009
#c79
#index119
#id122
#%
#!

#*highly adaptive distributed routing algorithm mobile wireless networks present new distributed routing protocol mobile wireless networks protocol family protocols term link algorithms protocol structured ordered sequence computations computation consisting sequence directed link protocol highly adaptive efficient scalable best suited use large dense mobile networks networks protocol link failures typically involves single distributed algorithm unique protocols face network partitions results protocol high degree desirable behavior achieved novel use physical establish temporal order topological change events used structure order algorithm topological changes refer protocol ordered routing algorithm 
#@Vincent D. Park, M Corson
#t2009
#c89
#index120
#id123
#%121,122,123,124,125,126,127,128,129,130,131,132,133,134,135
#!

#*considerations mobile mesh mobile communications central operations necessary communicate simultaneously multiple common task paper describes problem routing resource mobile mesh networks presents recommendations necessary internet protocols operate effectively environments mobile mesh network autonomous mobile connected wireless links form arbitrary graph free network wireless topology change rapidly 
#@M Corson, J Macker, S Batsell
#t2009
#c90
#index121
#id124
#%
#!

#*distributed routing protocol algorithm constructing maintaining routing tables communication networks presented algorithm employed message circuit networks uses distributed computation provides routing tables loop free destination times changes network completely means arbitrary failures network finite time sense providing routing paths connected nodes destination routes updated update destination 
#@P Merlin, A Segall
#t2009
#c91
#index122
#id125
#%
#!

#*distributed routing algorithm networks new distributed algorithm presented dynamically determining weighted shortest paths used message routing networks major features algorithm paths defined form weights change number steps required new shortest paths network links previous algorithms specifically worst case time largest number weighted shortest paths previous loop free distributed algorithms time 
#@Jeffrey M. Jaffe, Franklin H. Moss
#t2009
#c92
#index123
#id126
#%
#!

#*distributed routing labeled author presents new routing algorithm called labeled distance vector routing algorithm loop free problem distributed algorithm operates arbitrary link node provides shortest paths finite time occurrence arbitrary sequence topological changes contrast previous successful approaches loop free routing maintains independence property does require multiple new algorithm shown loop free converge finite time arbitrary sequence topological changes performance compared performance distributed routing algorithms 
#@J.J. Garcia-Luna-Aceves
#t2009
#c89
#index124
#id128
#%
#!

#*loop free routing using computations family distributed algorithms dynamic computation shortest paths network presented analyzed according algorithms node maintains vector distance node update messages node message contains vector entries entry length selected path network destination entry update query previous query new algorithms treat problem distributed shortest path routing computations proposed improve algorithms introduced previously author new algorithms shown converge finite time arbitrary sequence link topological changes loop free outperform loop free routing algorithms previously proposed 
#@J.J. Garcia-Luna-Aceves
#t2009
#c93
#index125
#id129
#%
#!

#*loop free path finding algorithm complexity path finding algorithm lpa presented lpa second hop distance destination ensure termination addition uses inter neighbor mechanism detailed lpa presented complexity evaluated lpa average performance compared simulation performance representative state art distributed routing ideal link state algorithm algorithm based multiple dual simulation results lpa scalable alternative dual terms average number steps messages operations needed algorithm converge topology change lpa shown achieve loop additional overhead prior algorithms based second hop distance information 
#@J.J. Garcia-Luna-Aceves, S. Murthy
#t2009
#c94
#index126
#id130
#%
#!

#*distributed scalable routing based vectors link states link vector algorithms distributed routing information networks according maintains subset topology links links used neighbor paths known based subset topology information paths state information 
#@J.J. Garcia-Luna-Aceves, Jochen Behrens
#t2009
#c95
#index127
#id131
#%
#!

#*distributed algorithms generating loop free routes networks frequently changing topology consider problem maintaining communication nodes data network central presence frequent topological changes example mobile packet radio networks argue flooding schemes significant networks propose class distributed algorithms new loop free routes node left route topology built redundancy algorithms typically involve communication network affected topological change 
#@MG ELI, P DIMITRI
#t2009
#c96
#index128
#id132
#%
#!

#*distributed routing algorithm mobile wireless networks present loop free distributed routing protocol mobile packet radio networks protocol intended use networks rate topological change fast make flooding possible routing method slow make existing protocols nearly static topology applicable routing algorithm distributed fashion arbitrary changes topology global topological knowledge protocol ability maintain source loop free routing desired minimal overhead varying topology protocol performance measured terms end end packet compared pure flooding alternative algorithm suited high rate topological change environment protocol emphasis examining performance measures vary function rate topological changes network topology message traffic level results indicate new protocol generally outperforms alternative protocol rates change traffic conditions true traffic protocols significantly outperform flooding rates change high rates algorithms network topology dense connected seen major factor relative performance algorithms 
#@M Corson, A. Ephremides
#t2009
#c97
#index129
#id133
#%
#!

#*highly dynamic distance vector routing mobile computers ad hoc network cooperative collection mobile hosts required access point paper present design operation ad hoc networks basic idea design operate mobile host specialized view interconnection topology mobile hosts network amounts new sort routing protocol 
#@Charles Perkins, Pravin Bhagwat
#t2009
#c98
#index130
#id134
#%
#!

#*efficient routing protocol wireless networks present wireless routing protocol routing nodes communicate distance second hop destination reduces number cases routing loop occur accounts fast convergence properties detailed presented performance compared simulation performance distributed algorithm dual loop free distance vector algorithm ideal link state algorithm represent state art internet routing simulation results indicate efficient alternatives analyzed 
#@S. Murthy, J.J. Garcia-Luna-Aceves
#t2009
#c99
#index131
#id135
#%
#!

#*dynamic source routing ad hoc wireless networks ad hoc network collection wireless mobile hosts network established environment mobile host hosts packet destination limited range mobile host wireless paper presents routing ad hoc networks uses dynamic source routing protocol quickly host movement frequent requires little overhead frequently based results packet level simulation mobile hosts operating ad hoc network protocol performs variety conditions movement rates highest rates host movement overhead protocol quite low just total data movement network mobile hosts cases difference length routes used route lengths cases route lengths average factor optimal 
#@David B. Johnson, David A. Maltz
#t2009
#c100
#index132
#id136
#%
#!

#*network time protocol network time protocol protocol systems packet variable data networks uses layer designed particularly effects variable 
#@DL Mills
#t2009
#c101
#index133
#id137
#%
#!

#*time ordering events distributed concept event distributed examined shown define partial ordering events distributed algorithm given used order events use total ordering illustrated method solving problems algorithm specialized physical bound derived far 
#@Leslie Lamport
#t2009
#c7
#index134
#id138
#%
#!

#*distributed routing algorithm mobile radio networks present distributed routing protocol mobile packet radio networks protocol intended use networks rate topological change fast make flooding possible routing method slow make existing protocols static topology applicable ing algorithm distributed fashion arbitrary changes topology global topological knowledge protocol maintains set loop free routes destination node route protocol performance measured terms end end packet compared pure flooding alternative algorithm suited topological change environment results rate topological changes high flooding alternatives lower rates change appears effects channel access account performance new algorithm generally superior alternative protocols 
#@M. Scott Corson, Anthony Ephremides
#t2009
#c97
#index135
#id139
#%
#!

#*lexicon based approach opinion mining important types information web opinions expressed user generated content customer reviews products posts blogs paper focus customer reviews products particular study problem determining semantic orientations positive negative opinions expressed product features reviews problem applications opinion mining summarization search existing techniques utilize list opinion words called opinion lexicon purpose opinion words words express desirable great bad poor states approaches major paper propose lexicon based approach solving problem exploiting external linguistic natural language expressions approach allows handle opinion words context dependent major difficulties existing algorithms special words phrases language opinions based linguistic patterns effective function multiple opinion words sentence called opinion observer based proposed technique implemented experimental results using benchmark product review data set additional reviews proposed technique highly effective outperforms existing methods significantly 
#@Xiaowen Ding, Bing Liu, Philip S. Yu
#t2009
#c102
#index136
#id140
#%137,138,139,140,141,142,143,144,145,146,147,148,149,150,151
#!

#*mining comparative sentences relations paper studies text mining problem comparative sentence mining comparative sentence ordering relation sets entities respect common features example comparative sentence better comparative relation better given set evaluative texts web reviews postings news articles task comparative sentence mining identify comparative sentences texts extract comparative relations identified comparative sentences problem applications example product manufacturer wants know customer opinions products comparison paper propose novel techniques based new types sequential rules perform tasks experimental evaluation conducted using different types evaluative texts web results techniques promising 
#@Nitin Jindal, Bing Liu
#t2009
#c103
#index137
#id141
#%
#!

#*automatic construction polarity tagged corpus html documents paper proposes novel method building polarity tagged corpus html documents characteristics method fully automatic applied arbitrary html documents idea method utilize certain layout structures linguistic pattern using automatically extract sentences express opinion experiment method construct corpus consisting sentences 
#@Nobuhiro Kaji, Masaru Kitsuregawa
#t2009
#c104
#index138
#id142
#%
#!

#*determining sentiment opinions identifying sentiments parts opinions challenging problem present given topic automatically finds people opinions topic sentiment opinion contains module determining word sentiment combining sentiments sentence experiment various models classifying combining sentiment word sentence levels promising results 
#@Soo-Min Kim
#t2009
#c105
#index139
#id143
#%
#!

#*automatic identification pro reasons online reviews paper present automatically pros cons online reviews approaches developed extracting opinions text focus extracting reasons opinions form fact opinion online review sites author generated pros cons propose pros cons sentences review texts maximum entropy model trained resulting labeled set extract pros cons online review sites explicitly provide experimental results resulting identifies pros cons precision recall 
#@Soo-Min Kim, Eduard H. Hovy
#t2009
#c104
#index140
#id144
#%
#!

#*opinion extraction summarization tracking news blog corpora humans like express opinions know opinions automatically mining organizing opinions heterogeneous information sources useful individuals organizations opinion extraction opinion summarization opinion tracking important techniques understanding opinions opinion extraction opinions word sentence document levels articles opinion summarization opinions articles sentiment degree correlated events paper news web blog articles investigated trec articles collected web blogs serve information sources opinion extraction documents related issue selected experimental algorithms opinion extraction word sentence document level proposed issue relevant sentence selection discussed topical opinionated information opinion representative sentences text based summaries different languages different sources compared finally opinionated showing degree illustrated opinion tracking 
#@Lun-Wei Ku, Yu-Ting Liang, Hsin-Hsi Chen
#t2009
#c106
#index141
#id145
#%
#!

#*opinion observer analyzing comparing opinions web web excellent source consumer opinions numerous web sites containing opinions customer reviews products discussion groups blogs paper focuses online customer reviews products makes proposes novel framework analyzing comparing consumer opinions products prototype called opinion observer implemented single visualization user able clearly product consumers terms various product features comparison useful potential customers product potential customer visual feature feature comparison consumer opinions products helps decide product buy product manufacturer comparison enables easily gather marketing intelligence product information second new technique based language pattern mining proposed extract product features pros cons particular type reviews features form basis comparison experimental results technique highly effective outperform existing methods significantly 
#@Bing Liu, Minqing Hu, Junsheng Cheng
#t2009
#c38
#index142
#id146
#%
#!

#*examining role linguistic knowledge sources automatic identification classification reviews paper examines problems document level sentiment analysis determining given document review classifying polarity review positive negative demonstrate review identification performed high accuracy using features examine role types simple linguistic knowledge sources polarity classification 
#@Vincent Ng, Sajib Dasgupta, S. M. Niaz Arifin
#t2009
#c104
#index143
#id147
#%
#!

#*exploiting class relationships sentiment categorization respect rating scales address rating inference problem simply decide review thumbs thumbs previous sentiment analysis work determine author evaluation respect multi point scale task represents interesting standard multi class text categorization different similarity class labels example intuitively evaluate human performance task apply meta algorithm based metric labeling formulation problem given classifier output explicit attempt ensure similar items similar labels meta algorithm provide significant improvements multi class regression versions svms employ novel similarity measure appropriate problem 
#@Bo Pang, Lillian Lee
#t2009
#c104
#index144
#id148
#%
#!

#*thumbs sentiment classification using machine learning techniques consider problem classifying documents topic overall sentiment determining review positive negative using movie reviews data standard machine learning techniques outperform human produced machine learning methods employed naive bayes maximum entropy classification support vector machines perform sentiment classification traditional topic based categorization conclude examining factors make sentiment classification problem challenging 
#@Bo Pang, Lillian Lee, Shivakumar Vaithyanathan
#t2009
#c107
#index145
#id149
#%
#!

#*extracting product features opinions reviews consumers line reviews order make informed product choice paper introduces unsupervised information extraction reviews order build model important product features evaluation relative quality products compared previous work achieves higher precision lower recall feature extraction task novel use labeling finding semantic orientation words context leads strong performance tasks finding opinion phrases polarity 
#@Ana-Maria Popescu, Oren Etzioni
#t2009
#c108
#index146
#id150
#%
#!

#*learning extraction patterns subjective expressions paper presents bootstrapping process learns rich extraction patterns subjective opinionated expressions high precision classifiers data automatically create large training set given extraction pattern learning algorithm learned patterns used identify subjective sentences bootstrapping process learns subjective patterns increases recall maintaining high precision 
#@Ellen Riloff, Ellen Riloff
#t2009
#c109
#index147
#id151
#%
#!

#*opinion summarization linking sources target problem linking source entity source coreference resolution needed creating opinion summaries paper source coreference resolution transformed standard noun phrase coreference resolution apply state art coreference resolution approach transformed data evaluate available corpus manually annotated opinions 
#@Veselin Stoyanov, Claire Cardie
#t2009
#c110
#index148
#id152
#%
#!

#*word sense subjectivity subjectivity meaning important properties language paper explores interaction empirical evidence support hypotheses subjectivity property associated word senses word sense disambiguation directly subjectivity 
#@Janyce Wiebe, Rada Mihalcea
#t2009
#c104
#index149
#id153
#%
#!

#*answering opinion questions separating opinions identifying polarity opinion sentences opinion question answering challenging task natural language processing paper discuss necessary component opinion question answering separating opinions fact document sentence level present bayesian classifier documents opinions news stories unsupervised statistical techniques significantly task detecting opinions sentence level present model classifying opinion sentences positive negative terms main perspective expressed opinion results large collection news stories human evaluation sentences reported achieve high performance document classification precision recall performance detecting opinions classifying sentence level positive negative accuracy 
#@Hong Yu, Vasileios Hatzivassiloglou
#t2009
#c111
#index150
#id154
#%
#!

#*movie review mining summarization web online review useful important information resource people result automatic review mining summarization research topic recently different traditional text summarization review mining summarization aims extracting features express opinions determining opinions positive negative paper focus specific domain movie review multi knowledge based approach proposed integrates wordnet statistical analysis movie knowledge experimental results effectiveness proposed approach movie review mining summarization 
#@Li Zhuang, Feng Jing, Xiaoyan Zhu
#t2009
#c32
#index151
#id155
#%
#!

#*data mining problem customer oriented catalog segmentation framework data mining assumes enterprise decision overall utility customers contribution customer function data available customer catalog segmentation enterprise wants design product catalogs size maximize overall number catalog products applications customer enterprise purchase products ones contained catalog paper investigate alternative problem formulation customer oriented catalog segmentation overall utility measured number customers specified minimum catalogs introduce customer oriented catalog segmentation problem discuss complexity investigate different design efficient approximate algorithms customer oriented catalog segmentation problem deterministic randomized algorithms algorithms local randomized algorithms depend reasonable initial solution explore combination experimental evaluation synthetic real data demonstrates new algorithms yield catalogs significantly higher utility compared classical catalog segmentation algorithms 
#@Martin Ester, Rong Ge, Wen Jin, Zengjian Hu
#t2009
#c3
#index152
#id156
#%153,154,155,156,157,158,159,160,161,162
#!

#*data mining framework optimal product selection retail data generalized model recent years data mining researchers developed efficient association rule algorithms retail market adopt association rules optimize retail marketing context previous paper authors introduced product selection model called model selects interesting products product based cross selling potential given 
#@Tom Brijs, Bart Goethals, Gilbert Swinnen, Koen Vanhoof, Geert Wets
#t2009
#c107
#index153
#id157
#%
#!

#*threshold set cover given subsets set cover problem selecting possible subsets covers max cover problem subsets maximum cardinality problems hard prove threshold set cover efficiently np time algorithms gap low order terms ratio approximation algorithm previous results showed ratio log max cover low order terms assumption categories subject descriptors algorithms problems discrete structures 
#@Uriel Feige
#t2009
#c112
#index154
#id158
#%
#!

#*view data mining present framework based optimization evaluating data associations clustering terms utility framework leads quickly interesting computational analysis segmentation theory science university fi email research 
#@Jon M. Kleinberg, Christos H. Papadimitriou, Prabhakar Raghavan
#t2009
#c113
#index155
#id159
#%
#!

#*mining values added association rules value added product term addition major products paper term semantic addition known association rules consider addition numerical values attribute values price profit degree level security lead notion random variables added value attributes data model probabilistic considerations data mining 
#@Rakesh Agrawal, Tomasz Imielinski, Arun Swami
#t2009
#c114
#index156
#id160
#%
#!

#*theoretical framework data mining research data mining knowledge discovery databases developing good algorithms various data mining tasks example recent parts research effort data mining process user interface issues database topics visualization relatively little published theoretical foundations data mining paper present possible theoretical approaches data mining area questions answers paper 
#@Heikki Mannila
#t2009
#c115
#index157
#id161
#%
#!

#*efficient algorithms creating product catalogs purposes paper define catalog catalog collection products items presented customer purchase single problem addresses build collection catalogs customers customer achieve optimal profit catalog subset given set items different catalogs contain items catalogs overlap general important extension single problem seeks optimal set catalogs multiple multiple catalogs customer catalog creation important applications commerce traditional especially used personalized recommender systems catalog creation problem np complete relatively expensive approximation algorithms recently developed paper efficient techniques building catalogs algorithms outperform previously suggested approaches techniques previously suggested feasible numbers customers catalogs techniques directly use objective function maximize profit optimal solution approach based gradient combining techniques clustering similar customers better results obtained analyze performance algorithms respect theoretical bound cases performance close optimal 
#@Michael Steinbach, G Karypis, Vipin Kumar
#t2009
#c116
#index158
#id162
#%
#!

#*profit item selection cross selling considerations literature data mining different algorithms association rule mining proposed relatively little study association rules specific paper applications association rules profit item selection cross selling effect problem investigated problem selecting subset items profit consideration cross selling prove simple version problem np hard propose new approach problem consideration loss rule association rule model cross selling effect problem transformed quadratic programming problem case quadratic programming applicable propose heuristic approach experiments conducted proposed methods highly effective efficient 
#@Raymond Chi-Wing Wong, Ada Wai-Chee Fu, Ke Wang
#t2009
#c14
#index159
#id163
#%11,12,163,153,164,165,166,167,168,169,155,170,171,172,173,174
#!

#*item selection authority profit ranking fundamental problem business applications ranking items respect notion profit based historical transactions difficulty profit item sales influence sales items cross selling effect paper draw influence mutual authority web pages based present novel approach item 
#@Ke Wang, Ming-Yen Thomas Su
#t2009
#c117
#index160
#id164
#%
#!

#*approximate catalog segmentation problem using programming consider catalog segmentation problem introduced fi problem given set items family subsets integer fi objective subsets subset sum max known simple algorithm performance 
#@Dachuan Xu, Yinyu Ye b, Jiawei Zhang c
#t2009
#c118
#index161
#id165
#%
#!

#*profit mining patterns actions major data mining applications gap based pattern extraction value based present profit mining approach reduce given set past transactions pre selected like build model target items strategies new customers goal identify issues profit mining propose 
#@Ke Wang, Senqiang Zhou, Jiawei Han
#t2009
#c0
#index162
#id166
#%
#!

#*heuristic algorithms binary quadratic programming problem paper consider binary quadratic programming problem quadratic objective suitable choice binary zero present heuristic algorithms based search computational results presented number available data sets involving interesting feature results problems search 
#@J.E. Beasley
#t2009
#c119
#index163
#id167
#%
#!

#*using association rules product case study discovery association rules suited applications market analysis purchase behaviour recent work indicates discovery interesting rules fact addressed study integrates discovery frequent model product selection model enable integration quantitative 
#@Tom Brijs, Gilbert Swinnen, Koen Vanhoof, Geert Wets
#t2009
#c3
#index164
#id169
#%
#!

#*computers guide theory np book theory np computational book features providing np complete problems updated book book does cover recent development theorem study search engine book reference science literature 
#@M. R. Garey, David S. Johnson
#t2009
#c50
#index165
#id170
#%
#!

#*mining frequent patterns candidate generation mining frequent patterns transaction databases time series databases kinds studied data mining research previous studies adopt set generation test approach candidate set generation costly especially exist large number patterns long patterns study propose pattern tree fp tree structure extended prefix storing compressed crucial information frequent patterns develop efficient fp tree based mining method fp growth mining complete set frequent patterns pattern growth efficiency mining achieved techniques large database compressed smaller data structure fp tree costly database fp tree based mining pattern growth method avoid costly generation large number candidate sets based method used mining task set smaller tasks patterns conditional databases dramatically reduces search space shows fp growth method efficient scalable mining long short frequent patterns order magnitude faster algorithm faster recently frequent pattern mining methods 
#@Jiawei Han, Jian Pei, Yiwen Yin
#t2009
#c10
#index166
#id171
#%
#!

#*finite algorithm maximize certain functions paper algorithm finds exact maximum certain nonlinear functions performing finite number operations objective functions need allow closed form solution sets equations order conditions associated transformed objective function examples especially homogeneous positive finite algorithm available methods algorithm uses internal representation feasible set varying dimensions linear programming gradient criterion used select sequence contain corresponding sequence increasing relative relatively interior shown global maximum feasible set interior finite way affine subsets vertices necessary requirement explicitly paper presents flow structure algorithm supporting theory decision theoretic use example computed version method 
#@B. von Hohenbalken
#t2009
#c120
#index167
#id172
#%
#!

#*introduction global optimization accurate modelling real world problems requires terms introduced model objective function constraints programming fields optimization presenting challenges practical theoretical aspects presence multiple local application global optimization techniques paper course global optimization techniques programming theoretical aspects nonlinear programming current state art algorithms global optimization follows examples problems general description phase algorithms local conditions short stochastic global example depth study deterministic spatial nlp advances programming theory constraints 
#@R Horst, PM Pardalos, NV Thoai
#t2009
#c121
#index168
#id173
#%
#!

#*quadratic binary programming approach determine seizures common focus normal brain areas results brain spatiotemporal states seizures brain sites index measures activity eeg brain applying optimization theory particular quadratic zero programming able select brain sites seizures seizures seizures seizures optimal selection sites procedure applied research time shows prediction seizures advance occurrence 
#@Leonidas D. Iasemidis, PM Pardalos, J. Chris Sackellares, Deng-Shan Shiau
#t2009
#c122
#index169
#id174
#%
#!

#*authoritative sources hyperlinked environment network structure hyperlinked environment rich source formation content environment provided effective means develop set tools extracting information link structures environments report experiments effectiveness variety contexts world wide web central address framework broad search topics discovery authoritative information sources topics propose formulation notion authority based set relevant authoritative pages set pages join link structure formulation connections certain matrices associated link graph connections turn heuristics link based analysis 
#@Jon M. Kleinberg
#t2009
#c112
#index170
#id175
#%
#!

#*methods problems data mining knowledge discovery databases data mining aim tools analysis large data consider methods used data mining search frequently occurring technique used various discuss data mining queries algorithms look use sampling data conclude open research problems data mining 
#@Heikki Mannila
#t2009
#c123
#index171
#id176
#%
#!

#*efficient algorithms discovering association rules association rules form relation value set fi introduced problem mining association rules large collections data method based database improved algorithm method based combinatorial analysis information obtained previous 
#@Heikki Mannila, Hannu Toivonen, A. Inkeri Verkamo
#t2009
#c124
#index172
#id177
#%
#!

#*optimizing web servers using page rank prefetching clustered accesses paper presents page rank based prefetching technique accesses web page clusters approach uses link structure requested page determine important linked pages identify page underlying approach case cluster accesses pages requested users web server typically based current previous pages requested furthermore requested pages links important page page higher probability requested experimental evaluation prefetching mechanism presented using real server logs results page rank based scheme does better random prefetching clustered accesses rates cases 
#@Victor Safronov, Manish Parashar
#t2009
#c125
#index173
#id178
#%
#!

#*computationally related problems look problems areas network theory artificial intelligence graph theory integer programming nonlinear programming related problems polynomial time present polynomial time algorithm problems known problems extend equivalence class problems known complete problem class languages polynomial time machines polynomial time deterministic machines related complete problems classes languages complete problem polynomial deterministic solution view appears likely equivalence class defines class problems solved deterministic polynomial time 
#@Sartaj Sahni
#t2009
#c126
#index174
#id179
#%
#!

#*sentiment aware model predicting sales high popularity weblogs blogs short present wealth information assessing general public sentiments opinions paper study problem mining sentiment information blogs investigate ways use information predicting product sales performance based analysis complex nature sentiments propose sentiment plsa plsa blog entry viewed document generated number hidden sentiment factors training plsa model blog data enables obtain summary sentiment information embedded blogs present sentiment aware model utilize sentiment information plsa predicting product sales performance extensive experiments conducted movie data set compare alternative models account sentiment information model different feature selection method experiments effectiveness proposed approach 
#@Yang Liu, Xiangji Huang, Aijun An, Xiaohui Yu
#t2009
#c127
#index175
#id181
#%176,177,178,179,180,181,182,183,184,185,186,142,187,188,189,144,145
#!

#*latent dirichlet allocation latent dirichlet allocation lda generative probabilistic model collections discrete data text corpora lda level hierarchical bayesian model item collection modeled finite mixture underlying set topics topic turn modeled mixture underlying set topic probabilities context text modeling topic probabilities provide explicit representation document present efficient approximate inference techniques based variational methods em algorithm empirical bayes parameter estimation report results document modeling text classification collaborative filtering comparing mixture model probabilistic lsi model 
#@David M. Blei, Andrew Y. Ng, Michael I. Jordan
#t2009
#c128
#index176
#id182
#%190,191,192,193,65,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210,83
#!

#*temporal analysis online news blogs previous work temporal analysis news items documents largely focused broad categorization small text collections region large scale temporal analysis online news media blogs presented analysis global news media coverage year period demonstrate using hierarchical database named entities provide results extremely analysis news items aggregate maps media attention particular world compared economic data analysis suggests best media attention 
#@Angelo Dalli
#t2009
#c38
#index177
#id183
#%
#!

#*media right using information estimate political orientation web documents paper introduces simple method estimating orientation online entities field particular information used estimate political orientation hypertext documents type orientation political orientation document degree traditionally left right estimating documents political orientation personalized information retrieval recommender systems application method uses simple probabilistic model estimate strength association document left right communities model estimates likelihood document small number documents known orientation model tested sets data web documents political weblogs non documents accuracy obtained model based classifiers statistically significant levels 
#@Miles Efron
#t2009
#c32
#index178
#id184
#%
#!

#*predictive power online increasing fraction global online form blogs web pages array new collaborative technologies point topics certain individual products sufficiently popular allow targeted online tracking flow topics based analysis million sales rank values books period postings blogs media web pages able draw interesting hand queries produce matching postings volume sales ranks second queries automatically generated cases sales rank difficult predict general predictors use online postings successfully predict sales rank 
#@Daniel Gruhl, Ramanathan V. Guha, Ravi Kumar, Jasmine Novak, Andrew Tomkins
#t2009
#c3
#index179
#id185
#%
#!

#*information blogspace study dynamics information propagation environments low overhead personal using large collection weblogs time example domain characterize model collection levels present topic propagation corpus notion long running topics consisting topics generated outside world events community second present propagation individual individual drawing theory model flow propose employ algorithm underlying propagation network sequence posts report results 
#@Daniel Gruhl, Ramanathan V. Guha, David Liben-Nowell, Andrew Tomkins
#t2009
#c38
#index180
#id186
#%
#!

#*probabilistic latent semantic analysis probabilistic latent semantic analysis novel statistical technique analysis occurrence data applications information retrieval natural language processing learning text related compared standard latent semantic analysis linear algebra performs singular value decomposition occurrence tables proposed method based mixture decomposition derived latent class results principled approach foundation order avoid propose widely applicable generalization maximum likelihood model approach yields substantial consistent improvements latent semantic analysis number 
#@Thomas Hofmann
#t2009
#c129
#index181
#id187
#%
#!

#*dynamic real time online functional models propose dynamic model price online key features model operates makes different previous approaches consider static models model different respect information price incorporated model based traditional notion price level incorporates dynamics form price sense incorporates key features dynamic environment online use novel functional data methodology allows measure include dynamic price characteristics illustrate model diverse set different book categories significantly higher prediction accuracy compared standard approaches 
#@Wolfgang Jank, Galit Shmueli, Shanshan Wang
#t2009
#c3
#index182
#id188
#%
#!

#*words traditional notion word meaning used natural language processing lexical meaning used relatively objective notion lexical meaning different subjective meaning aim come subjective aspects meaning expressed written texts value expressed paper explores structure wordnet lexical database used assess meaning particular construct measures based semantic differential technique 
#@Jaap Kamps, Maarten Marx
#t2009
#c130
#index183
#id189
#%
#!

#*evolution blogspace propose new tools address evolution hyperlinked corpora define time graphs extend traditional notion directed graph capturing link creation point phenomenon time second develop algorithms time dense community tracking notion community evolution develop tools context blogspace space weblogs blogs study involves approximately links blogs create time graph blogs automatic analysis internal time study evolution connected component structure community structure time graph blogspace transition behavior end rapidly past year just metrics scale metrics community structure expansion shows measures years link blogspace sources introduce concept randomized blogspace observe similar evolution component corresponding increase community structure having demonstrated formation communities time turn activity active communities extend recent work discover dense intra community link creation 
#@Ravi Kumar, Jasmine Novak, Prabhakar Raghavan, Andrew Tomkins
#t2009
#c131
#index184
#id190
#%
#!

#*structure evolution blogspace critical look million individual entries blogs activity patterns time 
#@Ravi Kumar, Jasmine Novak, Prabhakar Raghavan, Andrew Tomkins
#t2009
#c7
#index185
#id191
#%
#!

#*probabilistic model news event detection news event detection red defined discovery previously events historical news corpus contents time information news articles red focus contents news articles research works carried finding better time information paper directions based following characteristics news articles hand news articles events hand similar articles event appear news sources generative model news articles provides data environments perform red consideration characteristics propose probabilistic model incorporate content time information unified framework model gives new representations news articles news events furthermore based approach build interactive red provides additional functions present events 
#@Zhiwei Li, Bin Wang, Mingjing Li, Wei-Ying Ma
#t2009
#c35
#index186
#id192
#%
#!

#*probabilistic approach spatiotemporal theme pattern mining weblogs mining weblogs analyzing spatiotemporal patterns applications multiple domains paper define novel problem mining spatiotemporal theme patterns weblogs propose novel probabilistic approach model themes spatiotemporal theme patterns simultaneously proposed model spatiotemporal theme patterns extracting common themes weblogs generating theme life given location generating theme given time period evolution patterns discovered comparative analysis theme life theme experiments different data sets proposed approach discover interesting spatiotemporal theme patterns effectively proposed probabilistic model general used spatiotemporal text mining domain time location information 
#@Qiaozhu Mei, Chao Liu, Hang Su, ChengXiang Zhai
#t2009
#c38
#index187
#id193
#%
#!

#*mixture model contextual text mining contextual text mining concerned extracting topical themes text collection context information time location comparing analyzing variations themes different contexts topics document usually related context document analyzing topical themes context potentially reveal interesting theme patterns paper models proposed previous work propose new general probabilistic model contextual text mining cover existing models special cases specifically extend probabilistic latent semantic analysis plsa model context variables model context document proposed mixture model called contextual probabilistic latent semantic analysis model applied interesting mining tasks temporal text mining spatiotemporal text mining author topic analysis cross collection comparative analysis empirical experiments proposed mixture model discover themes contextual variations effectively 
#@Qiaozhu Mei, ChengXiang Zhai
#t2009
#c3
#index188
#id194
#%
#!

#*sentiment analysis using subjectivity summarization based minimum sentiment analysis seeks identify underlying text example application classifying movie review thumbs thumbs determine sentiment polarity propose novel machine learning method applies text categorization techniques just subjective portions document extracting portions implemented using efficient techniques finding minimum graphs greatly cross sentence contextual constraints 
#@Bo Pang, Lillian Lee
#t2009
#c104
#index189
#id195
#%
#!

#*variational bayesian framework graphical models paper presents novel practical framework bayesian model selection probabilistic graphical approach posterior distributions structures latent variables free form naturally incorporates conjugate large sample approximations generally needs 
#@H Attias
#t2009
#c132
#index190
#id196
#%
#!

#*modern information retrieval information retrieval ir years expansion web world wide web modern inexpensive graphical user interfaces storage devices result traditional ir quite date led introduction new ir books recently believe great need book approaches field complete way science perspective user perspective book effort gap useful course information retrieval course topic www pages digital version book complete contents table contents chapters available line version ordered directly 
#@Ricardo A. Baeza-Yates, Berthier A. Ribeiro-Neto
#t2009
#c50
#index191
#id197
#%
#!

#*modeling annotated data consider problem modeling annotated multiple types instance type serves description type image hierarchical probabilistic mixture models aim data latent dirichlet allocation latent variable model ective modeling joint distribution types conditional distribution annotation given primary type conduct experiments database images assessing performance terms held likelihood automatic annotation text based image retrieval 
#@David M. Blei, Michael I. Jordan
#t2009
#c35
#index192
#id198
#%190,364,176,662,663,664,665,666,667,263,259
#!

#*theory probability probability theory mathematics concerned analysis random phenomena central objects probability theory random variables stochastic processes events mathematical non deterministic events measured quantities single occurrences time random fashion individual random event times sequence random events exhibit certain statistical patterns studied predicted representative mathematical results describing patterns large numbers central limit theorem mathematical foundation statistics probability theory essential human involve quantitative analysis large sets data methods probability theory apply description complex systems given partial knowledge state statistical great discovery probabilistic nature physical phenomena scales described 
#@Harold Jeffreys
#t2009
#c30
#index193
#id199
#%
#!

#*bayesian methods categorical data bayesian methods given finite category sampling observations missing category generalization dirichlet family prior distributions closed sampling posterior predictive probabilities multiple functions closed form expressions developed case nested reported sets bayesian estimates computed easily relative frequencies effective computational methods given general case example involving used illustrate theory simple special case categorical missing data way table cross classified data sub distinguishing sub reported distinguishing sub assumptions process itself category probabilities sub report possible bayesian inference sub based likelihood function sub sub sub likelihood considered intractable bayesian conjugate prior inference develop bayesian conjugate theory recognizing complete functions functions posterior distributions resulting dirichlet prior distributions known generalized dirichlet distributions corresponding posterior density functions similar form likelihood family distributions closed sampling tractable various senses including convenient 
#@James M. Dickey, Jhy-Ming Jiang, Joseph B. Kadane
#t2009
#c133
#index194
#id200
#%
#!

#*bayesian data analysis incorporating new updated information second edition text bayesian data analysis practice theory describing perform statistical analysis bayesian perspective changes new edition include added material bayesian methods connected approaches focus added chapter computation topics examples additional chapters current models bayesian data analysis models generalized linear mixed models book text reference working life 
#@Andrew Gelman, John B. Carlin, Hal S. Stern, D. B. Rubin
#t2009
#c30
#index195
#id201
#%
#!

#*probabilistic approach semantic representation semantic networks produced human data properties easily spatial explore semantic representation probability words di erent contexts captures representation statistical properties large scale structure semantic humans 
#@Thomas L. Griffiths, Mark Steyvers
#t2009
#c134
#index196
#id202
#%
#!

#*overview text retrieval conference trec text retrieval conference trec held early people working groups goal conference research groups discuss work new large test collection large variety retrieval techniques reported including methods using automatic term weighting natural language techniques relevance feedback advanced pattern matching results run common evaluation groups able compare effectiveness different techniques discuss differences affected performance 
#@Donna Harman
#t2009
#c52
#index197
#id203
#%
#!

#*experimental comparison clustering methods examine methods clustering high dimensions paper perform experimental comparison batch clustering algorithms expectation maximization em algorithm version em algorithm means algorithm model based hierarchical clustering learn naive bayes models hidden node using discrete variable data sets real synthetic 
#@Marina Meila, D. Heckerman
#t2009
#c129
#index198
#id204
#%
#!

#*probabilistic latent semantic indexing probabilistic latent semantic indexing novel approach automated document indexing based statistical latent class model factor analysis data training corpus text documents generalization expectation maximization algorithm model able deal domain specific words contrast standard latent semantic indexing lsi singular value decomposition probabilistic variant statistical foundation defines proper generative data model retrieval experiments number test collections indicate substantial performance direct term matching methods lsi particular combination models different proven 
#@Thomas Hofmann
#t2009
#c35
#index199
#id205
#%65,410,411,412,181,413,414,415,416,83,417,418
#!

#*statistical methods speech recognition book important research mathematical foundations speech recognition focuses underlying statistical techniques hidden markov models decision trees expectation maximization algorithm information theoretic criteria maximum entropy probability estimation parameter data clustering smoothing probability distributions author goal present principles clearly setting advantages self organization real data enable reader apply techniques 
#@Frederick Jelinek
#t2009
#c30
#index200
#id206
#%
#!

#*making large scale svm learning practical training support vector machine svm leads quadratic optimization problem bound constraints linear constraint despite fact type problem issues considered svm particular large learning tasks training examples optimization techniques general quadratic programs quickly intractable memory time requirements 
#@Thorsten Joachims
#t2009
#c135
#index201
#id207
#%
#!

#*learning graphical models graphical models probability theory graph theory provide natural tool dealing problems occur applied mathematics engineering uncertainty complexity particular play increasingly important role design analysis machine learning algorithms fundamental idea graphical model notion complex built combining simpler parts probability theory serves parts combined consistent providing ways interface models data graph theory provides intuitively interface humans model highly sets variables data structure itself naturally design efficient general purpose algorithms book presents depth exploration issues related learning graphical model chapters tutorial chapters inference bayesian networks monte carlo methods et al variational methods learning bayesian networks chapters cover wide range topics current research 
#@Michael I. Jordan
#t2009
#c30
#index202
#id208
#%
#!

#*introduction variational methods graphical models paper presents tutorial introduction use variational methods inference graphical models bayesian networks markov random fields present number examples including database belief network machine hidden markov models run exact inference algorithms introduce exploit large numbers original graphical model inference efficient inference model provides bounds probabilities original model general framework generating variational transformations based finally return examples demonstrate variational algorithms formulated 
#@Michael I. Jordan, Zoubin Ghahramani, Tommi Jaakkola, Lawrence K. Saul
#t2009
#c23
#index203
#id209
#%
#!

#*general lower bounds based generated higher order article outline algorithm generate lower bounds exponential function principle arbitrary implemented generate necessary terms machine partition function leading lower bounds variational parameters optimized bounds order practical generated 
#@Martijn A. R. Leisink, Hilbert J. Kappen
#t2009
#c129
#index204
#id210
#%
#!

#*expectation propagation generative aspect model generative aspect model extension multinomial model text allows word probabilities vary documents 
#@Thomas Minka, J Lafferty
#t2009
#c136
#index205
#id211
#%
#!

#*using maximum entropy text classification paper proposes use maximum text probability distribution widely used variety language tasks language modeling speech tagging text underlying principle external knowledge prefer distributions distribution derived training data maximum entropy 
#@K Nigam, J Lafferty, A McCallum
#t2009
#c19
#index206
#id212
#%
#!

#*text classification labeled unlabeled documents using em paper shows accuracy learned text classifiers improved small number labeled training documents large unlabeled important text classification problems obtaining training expensive large quantities unlabeled documents available introduce algorithm learning labeled unlabeled documents based combination expectation maximization em naive bayes classifier algorithm classifier using available labeled documents labels new classifier using labels documents convergence basic em procedure works data model assumptions practice result present algorithm improve conditions weighting factor contribution data use multiple mixture components class experimental results obtained using text different real world tasks use unlabeled classification error 
#@K Nigam, A McCallum, Sebastian Thrun, Tom M. Mitchell
#t2009
#c23
#index207
#id213
#%
#!

#*latent semantic indexing probabilistic analysis latent semantic indexing lsi information retrieval technique based term document matrix empirical success prediction prove certain conditions lsi does capturing underlying semantics corpus achieves improved retrieval propose technique random projection way 
#@Christos H. Papadimitriou, Prabhakar Raghavan, Hisao Tamaki, Santosh Vempala
#t2009
#c137
#index208
#id214
#%
#!

#*probabilistic models unified collaborative content based recommendation sparse data environments recommender systems product community information target products researchers developed collaborative content based hybrid propose unified probabilistic framework merging collaborative content based extend aspect model incorporate way occurrence data users items item relative influence collaboration data versus content data 
#@Alexandrin Popescul, Lyle H. Ungar, DM Pennock, S Lawrence
#t2009
#c129
#index209
#id215
#%
#!

#*improving multi class text classification naive bayes numerous text documents available electronic available documents represent massive information easily value collection requires organization work organizing documents automated text accuracy understanding systems greatly paper seek advance understanding commonly 
#@Tommi Jaakkola, Jason D. M. Rennie, Jason D. M. Rennie
#t2009
#c138
#index210
#id216
#%
#!

#*analysis weblog comments access weblogs commercial services studies usually limited content weblog posts important aspect distinguishing weblogs web pages ability weblog readers posts directly posting comments paper present large scale study weblog comments relation posts using corpus comments estimate overall volume comments analyze relation weblog popularity patterns measure contribution content various aspects weblog access 
#@G Mishne, N Glance
#t2009
#c139
#index211
#id217
#%212,213,214,184,215,216,217,218,219,220,221
#!

#*analysis weblog weblogs important new components internet provide individual users easy way online views furthermore applications allow weblogs linked intended individual use practice weblogs increasingly appear facilitate distributed important implications use technology medium collaboration given special characteristics weblogs supporting applications suited range purposes require different forms paper analyze potential weblog technologies using framework technologies special attention structures dynamics weblogs naturally based initial analysis make number recommendations research apply technologies processes knowledge management 
#@A. de Moor, L. Efimova
#t2009
#c140
#index212
#id218
#%
#!

#*automated trend discovery weblogs past years weblogs new communication medium internet paper application data mining information extraction nlp algorithms discovering trends subset approximately weblogs lists key key phrases key public web site com addition maintain index weblog entries search index implemented trend search graphs normalized trend line time search query provides way estimate relative word given topics time 
#@N Glance, Matthew Hurst, Takashi Tomokiyo
#t2009
#c141
#index213
#id219
#%
#!

#*gap genre analysis weblogs weblogs blogs frequently modified web pages entries sequence genre internet communication widespread popularity characteristics systematically described paper presents results content analysis selected weblogs comparing empirically features corpus popular nature weblogs finding number blog authors blogs interactive oriented external events importance blogs forms self expression based profile generated empirical analysis consider likely blog genre respect forms digital communication internet today advance predictions long term 
#@Susan C. Herring, Lois Ann Scheidt, Sabrina Bonus, Elijah Wright
#t2009
#c142
#index214
#id220
#%
#!

#*structure authority weblog community weblog medium personal come new form social interaction web distributed completely connected topic communication set weblogs form social authors paper seeks understand social implications linking community drawing collected project past years social network analysis employed resulting social structure measures authority explored popularity measured public influence measured citation metrics evaluated respect authority references popular 
#@Cameron Marlow
#t2009
#c143
#index215
#id221
#%
#!

#*blog spam language model present approach detecting link spam common blog comments comparing language models used blog pages linked comments contrast link spam filtering approaches method requires training hard rule sets knowledge complete web connectivity preliminary experiments identification typical blog spam promising results 
#@G Mishne, David Carmel, Ronny Lempel
#t2009
#c144
#index216
#id222
#%
#!

#*learning class distributions aspects influence performance achieved classifier created machine learning aspects related difference numbers examples class difference large learning difficulties learn concept related class work discuss issues related learning class distributions relationship cost sensitive learning 
#@Maria Carolina Monard, Gustavo Batista
#t2009
#c145
#index217
#id223
#%
#!

#*robust metric opinion paper describes automated detecting polar expressions topic components approach nlp polar language extraction machine learning based topic classifier components making simple accurate assumption topical sentence contains polar language polar language topic subject matter evaluate components assumption corpus online consumer messages based components discuss measure overall sentiment particular topic expressed online messages different people propose use bayesian statistics form aggregate opinion metric metric introduced polarity topic facilitate statistically valid comparisons opinion multiple topics 
#@K Nigam, Matthew Hurst
#t2009
#c146
#index218
#id224
#%
#!

#*bayesian approach filtering mail growing problem mail internet examine methods automated construction filters messages user mail stream problem decision theoretic framework able make use probabilistic learning methods notion differential cost produce filters especially appropriate task appear 
#@Mehran Sahami, S. Dumais, D. Heckerman, Eric Horvitz
#t2009
#c147
#index219
#id225
#%
#!

#*machine learning automated text categorization automated categorization classification texts categories years increased availability documents digital form need organize research community approach problem based machine learning techniques general inductive process automatically classifier learning set documents characteristics categories advantages approach knowledge engineering approach consisting manual definition classifier domain experts good effectiveness considerable terms expert power different domains survey discusses main approaches text categorization machine learning paradigm discuss issues different problems document representation classifier construction classifier evaluation 
#@Fabrizio Sebastiani
#t2009
#c59
#index220
#id226
#%
#!

#*makes weblog weblog studying weblogs used know list known features weblog software important heart weblog weblog different news site software like draw experience developing using weblog software radio using competitive products type published weblogs conference work subsequent versions art market weblog software 
#@Dave Winer
#t2009
#c50
#index221
#id227
#%
#!

#*automated ranking database query results ranking relevant results query popular paradigm information retrieval discuss challenges investigate approaches enable ranking databases including known techniques information retrieval present results preliminary experiments 
#@Sanjay Agrawal, Surajit Chaudhuri, Gautam Das, Aristides Gionis
#t2009
#c148
#index222
#id228
#%223,222,191,224,225,226,227,228,229,230,231,232,233,234,235,236,237,238,239,240,241,242,243,244,245
#!

#*keyword based search relational databases internet search engines search paradigm traditional database management systems offer powerful query languages allow keyword based search paper discuss enables search relational databases implemented using commercial relational database web server allows users browser end outline challenges discuss implementation including results extensive experimental evaluation 
#@Sanjay Agrawal
#t2009
#c55
#index223
#id229
#%
#!

#*empirical analysis predictive algorithms collaborative filtering collaborative filtering recommender database user preferences additional topics products paper algorithms designed task based correlation coefficients vector based similarity statistical bayesian predictive accuracy set representative use basic classes evaluation 
#@John S. Breese, D. Heckerman, Carl Myers Kadie
#t2009
#c129
#index224
#id230
#%
#!

#*selection queries relational databases mapping strategies performance evaluation applications users specify target values certain attributes requiring exact matches values return instead result queries typically rank best match given attribute values paper study advantages limitations processing query single range query traditional relational database management rdbms process efficiently particular study determine range query evaluate query exploiting statistics available rdbms impact quality statistics retrieval efficiency resulting scheme report experimental evaluation mapping strategies real rdbms sql server experiments new techniques robust significantly efficient previously known strategies requiring sequential scan data sets 
#@Nicolas Bruno, Surajit Chaudhuri, Luis Gravano
#t2009
#c6
#index225
#id231
#%
#!

#*evaluating queries web accessible databases query web search engine usually consists list keywords search engine best pages query query model multimedia collections general relational data certain applications example consider relation information available including location price range overall rating user queries relation simply specify user location target price range return best terms combination proximity user closeness match target price range overall rating processing queries efficiently challenging number reasons critical web applications relation attributes available external web accessible form interfaces query potentially large set candidate objects paper study process queries efficiently setting attributes users specify target values external autonomous sources variety access interfaces present algorithms processing queries evaluate using synthetic real web accessible data 
#@Amelie Marian, Nicolas Bruno, Luis Gravano
#t2009
#c6
#index226
#id232
#%
#!

#*sql paper study simple sql extension enables query explicitly limit cardinality query result examine impact query optimization run time execution components relational dbms presenting approaches approach aggressive approach exploiting cardinality relational query plans results obtained empirical study conducted using db demonstrate sql extension illustrate approaches implementing 
#@Michael J. Carey, Donald Kossmann
#t2009
#c149
#index227
#id233
#%
#!

#*reducing distance sql query engine recent paper proposed adding cardinality query result explicitly query query demonstrated having showed extend based query db based large performance queries explicitly supported database engine paper present new 
#@Michael J. Carey, Donald Kossmann
#t2009
#c8
#index228
#id234
#%
#!

#*integration heterogeneous databases common domains using queries based textual similarity databases contain like course numbers personal names place names correspond entities real previous work integration heterogeneous databases assumed local appropriate global domain cases assumption does determining considered identical require detailed knowledge world purpose user query 
#@William W. Cohen
#t2009
#c10
#index229
#id235
#%
#!

#*providing database like access web using queries based textual similarity databases contain like course numbers personal names place names correspond entities real world previous work integration heterogeneous databases assumed local appropriate global domain assume instead names given natural language text propose logic database integration called reasons explicitly similarity local names measured using vector space model commonly adopted statistical information retrieval implemented data integration based used successfully integrate information web sites domains 
#@William W. Cohen
#t2009
#c10
#index230
#id236
#%
#!

#*fuzzy queries multimedia database systems essential differences multimedia databases objects images traditional database interesting new issues particular consider example multimedia database reasonable ask images similar fixed furthermore different ways obtaining information 
#@Ronald Fagin
#t2009
#c1
#index231
#id237
#%
#!

#*optimal aggregation algorithms middleware assume object database scores attributes example object color grade red shape grade attribute sorted list lists object grade attribute sorted grade highest grade object overall grade obtained combining attribute using fixed aggregation function combining rule average determine objects objects highest overall naive algorithm access object database grade attribute fagin given algorithm fagin algorithm efficient aggregation functions optimal high probability worst case analyze simple algorithm threshold algorithm ta optimal sense ta essentially optimal just aggregation functions just high probability worst case sense database unlike requires large size grow database size grows ta requires small constant size ta allows early yields precise sense approximate version answers types access sorted access middleware grade object sorted list list sequentially random access middleware grade object list step consider scenarios random access expensive relative sorted access provide algorithms essentially optimal cases science 
#@Ronald Fagin, Amnon Lotem, Moni Naor
#t2009
#c137
#index232
#id238
#%
#!

#*fast algorithm indexing data mining visualization traditional multimedia datasets promising idea fast searching traditional multimedia databases map objects points space using feature extraction functions provided domain expert use highly spatial access methods answer types queries including query example type range query 
#@Christos Faloutsos, King-Ip (David) Lin
#t2009
#c150
#index233
#id239
#%
#!

#*probabilistic framework vague queries information databases probabilistic learning model vague queries missing information databases described instead retrieving set answers approach yields ranking objects database response query using relevance user objects retrieved ranking actual query overall retrieval quality improved different kinds conditions vague queries notion vague predicates introduced based underlying probabilistic model missing attribute values easily addition corresponding applied combination standard predicates logic extending standard database systems missing data 
#@Norbert Fuhr
#t2009
#c8
#index234
#id240
#%
#!

#*probabilistic relational model integration ir databases paper probabilistic relational model presented combines relational algebra probabilistic retrieval based certain independence assumptions operators relational algebra probabilistic algebra generalization standard relational algebra furthermore special join operator implementing probabilistic retrieval proposed applied typical document databases queries ask documents object database addition implicit ranking objects provided case query probabilistic indexing uses probabilistic join operator proposed algebra intended standard interface combined database ir systems basis implementing user interfaces 
#@Norbert Fuhr
#t2009
#c35
#index235
#id241
#%
#!

#*preference sql design implementation current search engines fuzzy predicates defined complex problem search engines implemented standard sql sql does directly understand notion preference sql sql preference model based partial orders presented paper preference queries like soft selection built base preference types powerful operator combined sql programming style guarantees great programming preference sql does cient standard sql including high level implementation skyline optimal pre processor approach enables application 
#@Werner Kiesling, Gerhard Kostler
#t2009
#c8
#index236
#id242
#%
#!

#*foundations preferences database systems services new challenges database technology powerful flexible modeling technique complex preferences preference queries preferences soft constraints best possible match making propose partial order semantics preferences closely matches people variety natural preferences model construct complex preferences means various preference model key new called preference engineering preference algebra given best matches query model investigate complex preference queries simpler ones algorithms standard sql extended preferences presented paper believe model appropriate extend database technology effective support 
#@Werner Kiesling
#t2009
#c8
#index237
#id243
#%
#!

#*vague user interface relational databases vague queries specific query concerned data match vague query target concerned data close target conventional database systems handle vague queries directly users specific queries match data article describes called vague handle vague queries directly principal concept vague extension relational data model data metrics values domain problem implementing data different users different notion distance vague incorporates features enable itself individual views users 
#@Amihai Motro
#t2009
#c37
#index238
#id244
#%
#!

#*mining aviation safety data goal data analysis aviation safety simple improve safety path goal hard identify data mining methods applicable task data available analyzed focus interesting results answers questions based recent research project news number aviation safety work collect analyze safety related data number areas data mining techniques provide new tools perform analyses considered easily currently aviation safety collect analyze reports combination manual automated methods data analysis safety domain data mining methods aviation safety tools database query report generation process actual analysis tools help extract useful information data research project application data mining techniques aviation safety data help aviation safety analysis task effort led creation tool called aviation safety data mining paper describes research effort experience data mining aviation safety data learned 
#@Zohreh Nazeri, Eric Bloedorn, Paul Ostwald
#t2009
#c10
#index239
#id245
#%
#!

#*approach integrating query refinement sql applications require content based similarity retrieval techniques support retrieval paradigm database systems critical area user subjectivity important aspect queries objects relevant user depends query refinement used handle user subjectivity similarity search paper explores enhance database systems 
#@Michael Ortega-Binderberger, Kaushik Chakrabarti, Sharad Mehrotra
#t2009
#c0
#index240
#id246
#%
#!

#*content based image retrieval relevance feedback technology advances areas image processing information retrieval ir separately long time successful content based image retrieval systems require integration need develop integration mechanisms link image retrieval model text retrieval model established text retrieval techniques approaches image feature vectors domain weighted term vectors ir domain proposed paper furthermore relevance feedback technique ir domain used content based image retrieval demonstrate effectiveness experimental results image retrieval precision increases using proposed integration approach 
#@Yong Rui, T.Huang, Sharad Mehrotra
#t2009
#c151
#index241
#id247
#%
#!

#*index based search engine querying xml data relevance ranking query languages xml support boolean retrieval query result possibly subset xml elements entire documents satisfy search conditions query search paradigm works highly xml data collections electronic catalogs searching information open environments web large ranked retrieval appropriate query result rank list xml 
#@Anja Theobald, Gerhard Weikum
#t2009
#c0
#index242
#id249
#%
#!

#*improved heterogeneous distance functions instance based learning techniques typically handle continuous linear input values handle input attributes value difference metric designed reasonable distance values attribute values continuous attributes requiring discretization map continuous values values paper proposes new heterogeneous distance functions called heterogeneous value difference metric value difference metric value difference metric new distance designed handle applications attributes continuous attributes applications new distance metrics achieve higher classification average previous distance functions datasets attributes 
#@D. Randall Wilson, Tony R. Martinez
#t2009
#c107
#index243
#id250
#%
#!

#*using fagin algorithm merging ranked results multimedia middleware distributed multimedia information allows users access data different different data sources ranked various criteria fagin gives algorithm efficiently merging multiple ordered streams ranked results form new stream ordered combination ranks paper implementation fagin algorithm actual multimedia middleware including novel incremental version algorithm supports dynamic exploration data algorithm perform single multimedia server effective distributed environment limited set queries assumptions makes random access limit dramatically experience provides better understanding important algorithm open problem distributed multimedia information systems 
#@Edward L. Wimmers, Laura M. Haas, Mary Tork Roth, Christoph Braendli
#t2009
#c152
#index244
#id251
#%
#!

#*feedback adaptive loop content based retrieval methods currently exist relatively simple queries relevance feedback large methods vector spaces objects represented vectors spaces typically research solution propose novel method handle queries user provides weights examples learns concept similar objects method existing relevance feedback methods base metrics learning vector spaces spaces addition method example driven user feature selection main present novel way estimate object set able objects support algo shows exploit metric index ing structures support range queries search empirical results demonstrate method rapidly recall 
#@Leejay Wu, Christos Faloutsos, Katia P. Sycara, Terry R. Payne
#t2009
#c8
#index245
#id252
#%
#!

#*clustering documents words using bipartite spectral graph partitioning document clustering word clustering studied problems existing algorithms cluster documents words separately simultaneously paper present novel idea modeling document collection bipartite graph documents words using clustering problem bipartite graph partitioning problem solve partitioning problem use new spectral clustering algorithm uses second left right singular vectors word document matrix yield good spectral algorithm properties shown singular vectors solve real np complete graph problem present experimental results resulting clustering algorithm works practice 
#@Inderjit S. Dhillon
#t2009
#c3
#index246
#id253
#%247,248,249,250,251,252,253,254,255,95,165,256,257,117,258,259,260,261
#!

#*distributional clustering words text classification paper applies distributional clustering al document classication words groups based class labels associated word like latent semantic indexing feature space maintaining high document accuracy experimental results obtained real sets reduce feature orders magnitude better latent indexing al class based al feature selection markov based feature selection aggressive results improved classication classication clustering 
#@L. Douglas Baker, A McCallum
#t2009
#c35
#index247
#id254
#%
#!

#*hierarchical using partitioning propose unsupervised partitioning algorithm document data sets particular algorithm shows excellent scalability large data collections produces high quality clusters competitive clustering algorithm yields information significant words cluster words naturally hierarchical structure produced 
#@Daniel Boley
#t2009
#c83
#index248
#id255
#%
#!

#*document categorization query generation world wide web using present agent exploring documents world wide web based user heart agent unsupervised categorization set documents combined process generating new queries used search new related documents filtering resulting documents extract ones closely related starting document categories given present overall architecture 
#@Daniel Boley, Maria L. Gini, Robert Gross, Eui-Hong Han, Kyle Hastings, G Karypis, Vipin Kumar, Bamshad Mobasher, Jerome Moore
#t2009
#c153
#index249
#id256
#%
#!

#*cluster based approach construction importance successful operation information retrieval recognized techniques support automatic generation largely paper describes approach automatic generation global based discrimination value model appropriate clustering algorithm method implemented applied document collections preliminary results indicate method produces improvements retrieval performance test collections 
#@Carolyn J. Crouch
#t2009
#c35
#index250
#id257
#%
#!

#*gather cluster based approach browsing large document collections document clustering use main categories clustering slow corpora running time quadratic number documents second clustering improve argue problems arise used attempt improve conventional clustering 
#@Douglas R. Cutting, Jan O. Pedersen, David R. Karger, J. W. Tukey
#t2009
#c35
#index251
#id258
#%
#!

#*efficient clustering large document collections portion scientific data naturally text given large unlabeled document collection organize collection clusters related documentsfi using vector space model text data high dimensional sparse numerical data efficiently cluster large document paper present time memory technique entire clustering 
#@Inderjit S. Dhillon, James Fan, Yuqiang Guan
#t2009
#c154
#index252
#id259
#%
#!

#*concept large sparse text data using clustering unstructured text documents increasingly common available mining data sets represents major using words features text documents represented high dimensional sparse vectors dimensions sparsity paper study certain means algorithm clustering document algorithm outputs clusters concept vector 
#@Inderjit S. Dhillon, Dharmendra S. Modha
#t2009
#c23
#index253
#id260
#%
#!

#*lower bounds partitioning graphs let partition graph vertices subsets containing vertices let number edges vertices different subsets let largest matrix sum matrix graph matrix sum elements sum matrix zero theorem given shows effect maximum degree node limited shown right hand function computational studies ratio bound lower bound partition number random graphs having nodes 
#@W. E. Donath, A. J. Hoffman
#t2009
#c155
#index254
#id261
#%
#!

#*pattern classification pattern classification consists assigning entities described groups statistical problem consideration known probability achieved means bayes decision practice sub optimal classifier constructed data neural network approaches problem nearest neighbor models based assessing 
#@Richard O.Duda, Peter E.Hart, David G. Stork
#t2009
#c30
#index255
#id262
#%
#!

#*study document representations multidimensional scaling indexing terms project better means studying evaluating document representations indexes present study focused understanding ways individuals index terms goal using improving indexing systems indexing systems use matching descriptors documents descriptors information humans similarity meaning different descriptors terms similarity appears discrimination power available human judgments matching question data human judgments similarity indexing terms available machine searching search accuracy improved present levels similarity relations index term matching procedures increase search accuracy procedures used study use 
#@Katter, V. Robert
#t2009
#c50
#index256
#id263
#%
#!

#*self organizing maps self organizing maps data visualization technique reduce dimensions data use self organizing neural networks problem data visualization attempts solve humans simply high dimensional data techniques created help understand high dimensional data techniques reducing dimensions data presented course multi dimensional scaling way reducing dimensions producing map usually dimensions similarities data grouping similar data items reduce dimensions similarities just idea like example constructed like left hand lower right right hand 
#@Tom Germano
#t2009
#c30
#index257
#id264
#%
#!

#*efficient document clustering clustering increasing importance linear constant time clustering algorithms slow real time simple way speed clustering speed distance heart clustering study techniques improving cost distance lsi determine techniques speed clustering affect quality resulting 
#@Hinrich Schutze, Craig Silverstein
#t2009
#c35
#index258
#id265
#%
#!

#*normalized image segmentation propose novel approach solving grouping problem vision local features image data approach aims extracting global image treat image tion graph partitioning problem propose novel global criterion normalized cut ing graph normalized cut criterion measures total different groups total similarity groups cient computational technique based generalized problem used criterion applied approach static images results 
#@Jianbo Shi, Jitendra Malik
#t2009
#c156
#index259
#id266
#%
#!

#*impact similarity measures web page clustering clustering web documents enables semi certain types clustering method suitable similarity associated similarity measures proposed past systematic impact similarity metrics quality possibly popular cost qualitatively different 
#@Alexander Strehl, Er Strehl, J Ghosh, Raymond Mooney
#t2009
#c157
#index260
#id267
#%
#!

#*effectiveness efficiency clustering document retrieval ad hoc network cooperative collection mobile hosts required access point paper present design operation ad hoc networks basic idea design operate mobile host specialized view interconnection topology mobile hosts network amounts new sort routing protocol investigated basic routing mechanisms specified make suitable dynamic self starting network mechanism required users utilize ad hoc networks address previous use related poor properties algorithms face links resulting time dependent nature interconnection topology describing links mobile hosts finally ways basic network layer routing modified provide layer support ad hoc networks 
#@EM Voorhees
#t2009
#c45
#index261
#id269
#%
#!

#*combining language model sentiment analysis opinion paper describes blog opinion retrieval task year conduct experiments developed language model used retrieve related blog unit interactive knowledge adopted query retrieve blog unit include opinion introduce novel extracting technology extract text retrieved blog finally lexicon based method used document opinion 
#@Xiangwen Liao, Donglin Cao, Songbo Tan, Yue Liu, Guodong Ding
#t2009
#c159
#index262
#id270
#%263,264,265
#!

#*language modeling approach information retrieval models document indexing document retrieval extensively studied integration classes models goal researchers difficult problem argue indexing model suggests better indexing model help solve problem making parametric assumptions lead better retrieval performance furthermore making prior assumptions similarity documents instead propose approach retrieval based probabilistic language modeling estimate models document approach modeling non parametric integrates document indexing document retrieval single model advantage approach collection statistics used retrieval models model implemented model tested empirically approach significantly outperforms standard weighting different collections query sets 
#@Jay M. Ponte, W. Bruce Croft
#t2009
#c35
#index263
#id271
#%
#!

#*automatic query expansion using smart smart information retrieval project completely automatic approaches understanding retrieval large quantities text work trec performing runs routing ad hoc language environments major focus massive query expansion adding terms query terms come known relevant documents case routing just retrieved documents case ad hoc 
#@Chris Buckley, Gerard Salton, James Allan, Amit Singhal
#t2009
#c160
#index264
#id272
#%
#!

#*hierarchical dirichlet language model natural language engineering discuss hierarchical probabilistic model predictions similar popular language modelling procedure known smoothing number interesting differences smoothing probabilistic view problem point new directions language modelling ideas paper applicable problems modelling speech protein sequences 
#@D. J. C. Mackay, L. Peto
#t2009
#c161
#index265
#id273
#%
#!

#*constrained means clustering consider practical methods adding constraints means clustering algorithm order avoid local solutions clusters clusters having points observe phenomena applying means datasets number dimensions number desired clusters propose explicitly adding constraints underlying clustering optimization problem requiring cluster minimum number points investigate resulting cluster assignment step preliminary numerical tests real datasets indicate constrained approach poor local solutions producing better summary underlying data means clustering 
#@P. S. Bradley, K. P. Bennett, A. Demiriz
#t2009
#c162
#index266
#id274
#%267,268,269
#!

#*density based indexing approximate nearest neighbor queries consider problem performing nearest neighbor queries efficiently large high dimensional databases assuming database scan determine nearest neighbor entries study constructing index structure database traditional database indexing algorithms high dimensional data depending scheme arguments nearest neighbor queries 
#@K. P. Bennett, Usama Fayyad, Dan Geiger
#t2009
#c163
#index267
#id275
#%
#!

#*nonlinear programming substantially pages improved edition best selling nonlinear programming book focuses iterative algorithms constrained optimization large scale problems interface continuous discrete optimization 
#@D.P. Bertsekas 1999
#t2009
#c50
#index268
#id276
#%
#!

#*compressed data aggregate query approximation continuous dimensions efficiently answering decision support queries work direction data cube queries efficiently pre parts cube having large space requirements pre computation requires hierarchy fixed dimensions categorical pre queries advantage pre computation fixed hierarchy existing pre computation techniques measure aggregation function cube queries target measure different aggregation functions require pre data paper propose new data cube reduces does require discretization hierarchy query dimension fixed potential target measure supports functions additional storage costs approximate relatively accurate answers queries mechanisms reduce error approximation evaluation indicates compression addresses limitations existing approaches 
#@Jayavel Shanmugasundaram, Usama Fayyad, P. S. Bradley
#t2009
#c3
#index269
#id277
#%
#!

#*constraint driven clustering clustering methods data driven need driven data driven methods discover true structure underlying data need driven methods aims organizing true structure certain application requirements need driven constrained clustering able useful clusters applications energy aware sensor networks privacy market segmentation existing methods constrained clustering require users provide number clusters unknown advance crucial impact clustering result paper argue natural way generate clusters let application specific constraints decide number clusters purpose introduce novel cluster model constraint driven clustering finds number compact clusters satisfy user provided constraints general types constraints considered minimum significance constraints minimum variance constraints types prove np problem different constraints propose novel dynamic data structure cd tree data points nodes node approximately constraints minimizes objective function based cd trees develop efficient algorithm solve new clustering problem experimental evaluation synthetic real datasets demonstrates quality generated clusters scalability algorithm categories subject descriptors 
#@Rong Ge, Martin Ester, Wen Jin, Ian Davidson
#t2009
#c3
#index270
#id278
#%0,271,272,273,274,275,266,276,277,278,279,152,280,281,282,283,284,155,285,286,287,288,289,290,291,292,293,294
#!

#*approximation algorithms anonymity consider problem table containing personal records individual privacy maintaining data possible techniques proposed literature release considered information corresponding individual release individuals information appears release order achieve entries table generalized age value age range goal little information possible release optimization problem anonymity problem anonymity problem np hard attribute values entries positive provide approximation algorithm problem improved positive results interesting cases specific values particular approximation algorithm special case anonymity approximation algorithm anonymity 
#@Gagan Aggarwal, Tomas Feder, Krishnaram Kenthapadi, Rajeev Motwani, Rina Panigrahy, Dilys Thomas, An Zhu
#t2009
#c164
#index271
#id279
#%
#!

#*max cluster formation wireless ad hoc networks ad hoc network represented set clusters clusterheads form hop dominating set node clusterheads form virtual backbone used route nodes cluster previous heuristics restricted hop clusters minimum hop dominating set problem np complete present heuristic form clusters wireless ad hoc network nodes assumed non deterministic pattern clusters formed node wireless links heuristic node wireless away value parameter heuristic heuristic run network changes features heuristic tends existing clusterheads network changes helps reduce communication transition clusterheads new clusterheads mobile nodes clusterheads clusterheads nodes heuristic simulation experiments demonstrate proposed heuristic better earlier heuristics degree based solutions 
#@Alan D. Amis, Ravi Prakash, Dung Huynh, Thai Vuong
#t2009
#c89
#index272
#id280
#%295,296,297,298,299,300,301,302,303,120,304
#!

#*energy efficient hierarchical clustering algorithm wireless sensor networks wireless network consisting large number small sensors low power effective tool data variety environments data collected sensor network single processing center uses reported data determine characteristics environment detect event communication message process designed limited energy resources sensors clustering sensors groups sensors communicate information clusterheads clusterheads communicate information processing center energy paper propose distributed randomized clustering algorithm organize sensors wireless sensor network clusters extend algorithm generate hierarchy clusterheads observe energy increase number levels hierarchy results stochastic geometry used derive solutions values parameters algorithm minimize total energy network sensors report data clusterheads processing center 
#@Seema Bandyopadhyay, Edward J. Coyle
#t2009
#c165
#index273
#id281
#%
#!

#*scalable clustering algorithms balancing constraints clustering methods data mining problems extremely scalable addition data mining applications clusters obtained balanced approximately size importance paper propose general framework scalable balanced clustering data clustering process steps sampling small representative subset points clustering data initial clusters data followed simple uniform sampling original data sufficient representative subset high probability proposed framework allows large class algorithms used clustering set focus popular parametric algorithms present algorithms clusters algorithm clusters based generalization problem refinement algorithm constrained iterative scheme complexity overall method log obtaining balanced clusters data points compares existing techniques balanced clustering addition providing balancing guarantees clustering performance obtained using proposed framework comparable better corresponding solution experimental results datasets including high dimensional ones provided demonstrate proposed framework 
#@A Banerjee, J Ghosh
#t2009
#c113
#index274
#id282
#%
#!

#*clustering scheme hierarchical control multi hop wireless networks paper present clustering scheme create hierarchical control structure multi hop wireless networks cluster defined subset vertices graph connected addition cluster required certain constraints useful management scalability hierarchy constraints simultaneously general graphs clustering obtained wireless network finally present efficient distributed implementation clustering algorithm set wireless nodes create set desired clusters 
#@Suman Banerjee, Samir Khuller
#t2009
#c89
#index275
#id283
#%
#!

#*minimum energy ad hoc networks minimum energy problem node adjust power order minimize total energy enable message source node nodes ad hoc wireless existing solutions node requires global network information including nodes network order decide paper new protocol 
#@Julien Cartigny, David Simplot, Ivan Stojmenovic
#t2009
#c89
#index276
#id284
#%
#!

#*clustering constraints feasibility issues means algorithm recent work extending means incorporate background information instance level link link introduce ways information form operate instances constraints easy complexity results feasibility 
#@Ian Davidson, S. S. Ravi
#t2009
#c166
#index277
#id285
#%
#!

#*identifying generating easy sets constraints clustering clustering constraints recent intelligence community recent work shown negative forms constraints associated sub problem just finding feasible clustering np complete worst case results entire problem class ing easy problem instances work large classes clustering constraints using easy sets constraints yields better illustrate sufficient conditions theory identify easy problem present algorithms create large easy constraint sets 
#@Ian Davidson, S. S. Ravi
#t2009
#c103
#index278
#id286
#%
#!

#*practical data oriented statistical disclosure control statistical disclosure control technique microdata statistical databases microdata individual records data vectors small prior aggregate contain data vectors disclosure individual information constant value data exact polynomial algorithms known date minimal variability loss methods literature rank data partition groups fixed size multivariate case ranking performed data vectors single paper candidate optimal solutions multivariate problems characterized case heuristics based hierarchical clustering genetic algorithms introduced data oriented natural data multivariate case fixed size hierarchical clustering algorithms presented require data projected single dimension methods clearly reduce variability loss compared conventional multivariate projected data 
#@Josep Domingo-Ferrer, Josep Maria Mateo-Sanz
#t2009
#c15
#index279
#id288
#%
#!

#*based approach data summarization privacy data summarization recognized fundamental operation database systems data mining important applications data compression privacy existing methods perform provide guarantees quality results paper introduce summarization approach numerical data based notion quality objective minimal set satisfying significance constraint given dataset proposed problem np complete design different approximation algorithms algorithms quality scale large databases approximation algorithms allows precise heuristic algorithm heuristic efficient algorithm multi dimensional index structures integrated database systems experiments heuristic algorithm summaries outperform state art data terms internal measures terms external measures using data summaries input clustering methods 
#@Rong Ge, Martin Ester, Wen Jin, Zengjian Hu
#t2009
#c167
#index280
#id289
#%
#!

#*optimal energy aware clustering sensor network sensor networks growing technologies potential changing collaborative dynamic distributed computing systems self organizing capabilities task efficient computation challenges implementation systems energy clustering order maintain certain degree service quality reasonable energy needs optimized stage operation sensor node clustering important optimization problem nodes clustered easily able communicate considering energy optimization parameter clustering paper study theoretical aspects clustering problem sensor networks application energy optimization illustrate optimal algorithm clustering sensor nodes cluster balanced total distance sensor nodes nodes balancing clusters needed load nodes total distance helps reducing communication overhead energy problem balanced clustering modeled flow problem solved using existing techniques 
#@Soheil Ghiasi, Ankur Srivastava, Xiaojian Yang, Majid Sarrafzadeh
#t2009
#c168
#index281
#id290
#%
#!

#*clustering visualization retail market transaction analysis including clustering market key application data mining retail domain specific requirements need obtaining easily results challenging characteristics stemming fact data thousands features highly non gaussian sparse chapter proposes relationship based approach clustering data dimensionality issue working suitable similarity space instead original high dimensional feature space similarity space satisfy business criteria requiring customer clusters represent comparable amounts apply efficient scalable graph partitioning based clustering techniques space output clustering algorithm used data points resulting similarity matrix dimensions clusters showing visualization assessing improving clustering example recommendations merging clusters easily derived user suitable number clusters results presented real retail data set customers products 
#@J Ghosh, Alexander Strehl
#t2009
#c30
#index282
#id291
#%
#!

#*joint optimization customer segmentation marketing policy maximize long term marketing media targeted direct mail internet marketing develop targeted marketing enhanced way feasible support substantially larger number marketing segments problem segments actions different segments increases substantially environment systematic procedure optimizing steps study present joint optimization approach issues segmentation customers homogeneous groups customers determining optimal policy set available actions segment implement joint optimization framework direct mail setting organization previous studies area importance following variables value use variables segment customers second step determine marketing policy optimal using markov decision processes following similar previous applications stochastic dynamic programming procedure based long run maximization expected average profit contribution combination steps optimization framework obtain optimal allocation marketing control segment policy performance procedure framework illustrated real life application results proposed model outperforms segmentation 
#@Jedid-Jah Jonker, Nanda Piersma, Dirk Van den Poel
#t2009
#c169
#index283
#id292
#%
#!

#*robust effective anonymity large databases privacy preserving data mining privacy requirements discovering original interesting patterns structures existing methods correlations attributes different attributes minimum level required policies paper propose novel privacy preserving transformation framework distance based mining operations based concept privacy preserving satisfy privacy constraint significance constraint framework robustness state art anonymity model privacy constraint minimum effectiveness significance constraint minimum number corresponding data records privacy preserving public data mining purposes original data records private present efficient methods generating maintaining privacy preserving data mining operations clustering easily public data represented instead private data records experiment demonstrates proposed methods achieve accurate results preserving privacy 
#@Wen Jin, Rong Ge, Weining Qian
#t2009
#c170
#index284
#id293
#%
#!

#*efficient clustering algorithms self organizing wireless sensor networks self organization wireless sensor networks involves network decomposition connected clusters challenging task limited bandwidth energy resources available paper make improving efficiency self organization wireless sensor present novel approach message efficient clustering nodes local growth introduce algorithms 
#@Rajesh Krishnan, David Starobinski
#t2009
#c171
#index285
#id294
#%
#!

#*complexity optimal anonymity technique proposed literature alternative way release public formation data privacy data prove general versions relations np hard ing version amounts choosing minimum number entries present polynomial time algorithm optimal anonymity achieves approximation ratio size database particular log approximation constant algorithm exponential algorithm condition log approximation degree believe algorithm potentially quite fast 
#@Adam Meyerson, Ryan Williams
#t2009
#c1
#index286
#id295
#%
#!

#*data provide anonymity information information internet access fast computers large storage increased volume information collected individuals data sources makes identify individuals private information released data time increasing organizations release data aggregate statistical information explicit number data characteristic data term exist allow data recipient identify individuals data refer paper provide computational disclosure technique information private table identity individual released data refer recognized approach linking data based concepts generalization stored values semantically consistent precise alternatives anonymity table provide anonymity contained data allow recipient released information set individuals smaller introduce generalized table minimal generalization table respect anonymity requirement optimization problem objective data providing protection algorithm given table efficiently minimal generalization provide anonymity 
#@Pierangela Samarati, Latanya Sweeney
#t2009
#c172
#index287
#id296
#%
#!

#*scalable approach balanced high dimensional clustering market paper presents novel similarity based clustering approach based constrained weighted graph particularly real life market characterized high dimensional highly sparse customer product matrices positive attribute values significant built known highly efficient graph partitioning algorithm scalable easily 
#@Alexander Strehl, J Ghosh
#t2009
#c173
#index288
#id297
#%
#!

#*anonymity model privacy consider data holder held collection person specific field structured data data holder wants share version data data holder release version private data scientific guarantees individuals subjects data identified data practically useful solution provided paper includes formal protection model named anonymity set policies release provides anonymity protection information person contained release individuals information appears paper examines identification anonymity policies anonymity protection model important forms basis real world systems known similar provide guarantees privacy 
#@Latanya Sweeney
#t2009
#c174
#index289
#id298
#%
#!

#*constraint based clustering large databases constrained clustering finding clusters satisfy user specified constraints highly desirable applications paper introduce constrained clustering problem traditional clustering algorithms means handle scalable constraint clustering algorithm developed study finding initial solution user specified constraints solution performing object constraints algorithm consists movement resolution finding optimal solution np hard propose heuristics algorithm scale large data sets using heuristic cluster sharing experiments effectiveness efficiency heuristics 
#@Anthony K. H. Tung, Raymond T. Ng, Laks V. S. Lakshmanan, Jiawei Han
#t2009
#c123
#index290
#id299
#%
#!

#*clustering instance level constraints clustering algorithms conduct search space possible organizations data set paper propose types instance level clustering constraints link link constraints incorporated clustering algorithm search data sets tested results indicate constraints increase clustering accuracy investigate relative type constraint type accuracy improvements depends behavior clustering algorithm constraints 
#@Kiri Wagstaff, Claire Cardie
#t2009
#c175
#index291
#id300
#%
#!

#*drawing planar graphs book presents important fundamental algorithms planar graph drawing easy understand extensively illustrated end chapter suitable use advanced level algorithms graph theory graph drawing information visualization computational geometry book serve useful reference source researchers field graph drawing software information visualization design 
#@Rene Weiskircher
#t2009
#c176
#index292
#id301
#%
#!

#*efficient data clustering method large databases finding useful patterns large datasets recently problems area identification clusters regions multi dimensional work does address problem paper presents data clustering method balanced iterative reducing clustering demonstrates 
#@Tian Zhang, Raghu Ramakrishnan, Miron Livny
#t2009
#c10
#index293
#id302
#%
#!

#*scalable balanced model based clustering paper presents general framework generative model based clustering algorithm provide balanced solutions clusters comparable sizes model based clustering algorithms viewed iterative step optimization process iterative model estimation sample assignment instead maximum likelihood assignment approach used sample assignment step efficient iterative heuristic developed reduce computational complexity step make balanced sample assignment algorithm scalable large datasets demonstrate approach clustering complex data spatial data high dimensional text documents eeg time series keywords model based clustering scalable algorithms balanced clustering constrained clustering 
#@Shi Zhong, J Ghosh
#t2009
#c166
#index294
#id303
#%305,306,307,308,266,246,309,310,311,312,313,314,89,315,316,317,318,319,320,321,288,322,323,260,293,324,325,326
#!

#*organization mobile radio network distributed algorithm paper consider problem organizing set mobile radio nodes connected network require reliable structure acquired face arbitrary topological changes node require structure achieved use central propose develop self starting distributed algorithm maintains connected architecture algorithm especially suited needs intra task communication network discussed paper 
#@D.J. Baker, A. Ephremides
#t2009
#c177
#index295
#id304
#%
#!

#*better heuristic orthogonal graph drawing orthogonal drawing graph edges sequences segments present linear time space algorithm draw connected graph grid size theta particular non planar non planar graphs improvement algorithm simple easy implement planar non planar 
#@T Biedl, G Kant
#t2009
#c178
#index296
#id305
#%
#!

#*routing ad hoc networks using minimum connected dominating sets virtual backbone structure ad hoc network order support fault routing ad hoc network virtual backbone backbone networks key ways change nodes used primarily routing computing updating routes primary routes computed shortest paths computation virtual backbone necessary provide routes handle failures dynamic nature virtual backbone approach routing problem levels update virtual backbone update routes key contribution paper alternatives finding updating virtual backbone virtual backbone small possible use approximation minimum connected dominating set ad hoc network topology virtual backbone hosts maintain local global topology network shortest paths pairs nodes 
#@Bevan Das, Vaduvur Bharghavan
#t2009
#c179
#index297
#id306
#%
#!

#*design concept reliable mobile radio networks frequency new architecture mobile radio networks called linked cluster architecture described methods implementing architecture using distributed control techniques presented illustrate fully distributed control methods combined hierarchical control create network robust respect node loss connectivity changes distributed algorithms presented deal formation linkage clusters activation network links study performance network algorithms simulation model developed use simulation construct software simulation tools illustrated simulation results shown example high frequency communication network 
#@A. Ephremides, Anthony
#t2009
#c19
#index298
#id307
#%
#!

#*distributed algorithms generating routes networks frequently changing topology consider problem maintaining communication nodes data network central presence frequent topological changes example mobile packet radio networks argue flooding schemes significant networks propose class distributed algorithms new loop free routes node left route topology built redundancy algorithms typically involve communication network affected topological change 
#@Eli M. Gafni, Student Member, Dimitri, P. Bertsekas, Senior Member
#t2009
#c180
#index299
#id309
#%
#!

#*mobile multimedia radio network paper present multi cluster multi hop packet radio network addresses challenges required mobile multi channel code multi hop environment topology handle routing bandwidth designed various types traffic node clustering channel access control underlying features architecture 
#@Mario Gerla, Jack Tzu-Chieh Tsai
#t2009
#c97
#index300
#id310
#%
#!

#*approximation algorithms combinatorial problems simple polynomial time heuristic algorithms finding approximate solutions various polynomial complete optimization problems analyzed respect worst case behavior measured ratio worst solution value chosen algorithm optimal value certain problems simple form problem optimization problem based testing algorithms ratio bounded constant independent problem size number set problems simple algorithms yield worst case grow log problem size problem finding maximum graph algorithm ratio does grow fast problem size depends algorithm 
#@David S. Johnson
#t2009
#c137
#index301
#id311
#%
#!

#*spatial reuse packet radio networks packet radio networks present challenging problems network communication channel shared network users critical resource order make efficient use shared resource variety channel access protocols sharing investigated sharing occur domains frequency time space paper concerned sharing channel reuse spatial domain survey results approaches topological design associated channel access protocols attempt optimize performance spatial reuse communication channel presented 
#@L. Kleinrock, J. Silvester
#t2009
#c181
#index302
#id312
#%
#!

#*selecting ad hoc wireless networks personal digital important enable communication large networks mobile framework long way need clear necessary support architectures fixed base routing ad hoc network supported devices network essential range quite limited ir links supporting fi fi wireless local area recognized ad hoc architectures important component paper model network nodes graph processors communicate edges processors know size network start topological goal select small subset nodes graph ad hoc clearly set node node subset graph theoretic called dominating finding dominating set graph known np present fast distributed algorithm finds dominating set size network nodes links provably close minimum cardinality 
#@AK Parekh
#t2009
#c182
#index303
#id313
#%
#!

#*considerations problem interconnection pattern circuit dimensional surface minimal area discussed natural patterns graphs containing connections require area order achieve compact restricted classes graphs considered example arbitrary trees bounded degree embedded linear area edges planar graphs embedded efficiently 
#@Leslie G. Valiant
#t2009
#c183
#index304
#id314
#%
#!

#*competitive learning algorithms vector quantization efficient representation encoding signals limited resources finite storage restricted bandwidth fundamental problem technical information processing systems typically encoding communication messages deal different sources noise paper propose unifying approach data compression robust vector quantization explicitly channel noise bandwidth limitations random elimination resulting algorithm able limit noise general communication addition presented model allows derive novel competitive neural networks algorithm covers topology preserving feature maps called neural algorithm maximum entropy soft max rule special cases furthermore methods based noise models improve design reducing local application novel robust vector quantization algorithm image compression 
#@Stanley C. Ahalt, Ashok K. Krishnamurthy, Prakoon Chen, Douglas E. Melton
#t2009
#c184
#index305
#id315
#%
#!

#*frequency sensitive competitive learning clustering high dimensional paper competitive learning mechanisms principles obtain clusters comparable sizes inputs normalized mechanisms effective achieving balanced grouping inputs high dimensional spaces illustrated experimental results clustering popular text data sets dimensional spaces respectively 
#@A Banerjee, J Ghosh
#t2009
#c184
#index306
#id316
#%
#!

#*model based gaussian non gaussian clustering classification maximum likelihood approach sufficiently general current clustering algorithms including based sum criterion criterion currently implemented does allow features orientation size shape common clusters clusters restricted gaussian distributions does allow noise propose ways limitations matrix allows specify features clusters practical framework non gaussian clustering means incorporating noise form poisson process described approximate bayesian method choosing number clusters given performance proposed methods studied simulation results methods applied analysis data set arising study results better previous analyses keywords include bayes factors classification hierarchical iterative mixture models 
#@Jeffrey D. Banfield, Adrian E. Raftery
#t2009
#c185
#index307
#id317
#%
#!

#*tutorial em algorithm application parameter estimation gaussian mixture hidden markov models maximum likelihood parameter estimation problem em algorithm used form em algorithm given develop em procedure applications finding parameters mixture finding parameters hidden markov model hmm baum algorithm discrete 
#@J. Bilmes
#t2009
#c50
#index308
#id318
#%
#!

#*cure efficient clustering algorithm large databases clustering data mining useful discovering groups identifying interesting distributions underlying data traditional clustering algorithms clusters similar sizes presence outliers propose new cluster ing algorithm called cure robust outliers identifies clusters having non wide size cure achieves representing cluster certain fixed number points selecting points cluster center cluster specified fraction having point cluster allows cure adjust geometry non helps effects outliers handle large databases cure employs combination random sampling random sample data set partitioned partition clustered partial clusters clustered second yield desired clusters experimental results quality clusters produced cure better existing algorithms furthermore demonstrate random sampling partitioning enable cure outperform existing algorithms scale large databases cluster ing quality 
#@Sudipto Guha, Rajeev Rastogi, Kyuseok Shim
#t2009
#c5
#index309
#id319
#%
#!

#*time approximation scheme clustering metric spaces metric clustering problem defined follows given metric partition sets order minimize value paper approximation scheme problem 
#@Piotr Indyk
#t2009
#c186
#index310
#id320
#%
#!

#*generalized clustering supervised learning data assignment clustering algorithms increasingly analyzing data considerable work effective increasingly specific contrast developed diverse clustering systematic framework views clustering process iterative optimization includes supervised learning instance suggested 
#@Annaka Kalton, Pat Langley, Kiri Wagstaff, Jungsoon P. Yoo
#t2009
#c3
#index311
#id321
#%
#!

#*good bad spectral develop natural measure assessing quality clustering existing measures simple recursive heuristic shown guarantees main result article analysis popular spectral algorithm variant spectral clustering case guarantees finds good clustering 
#@Ravi Kannan, Santosh Vempala, Adrian Vetta
#t2009
#c187
#index312
#id322
#%
#!

#*clustering algorithms data meaningful useful groups called clusters intra cluster similarity inter cluster similarity discovered clusters used explain characteristics underlying data distribution serve foundation various data mining analysis techniques applications clustering include different customer groups based patterns categorization documents world wide web grouping similar grouping spatial locations data software clustering low high dimensional datasets analyzing characteristics various clusters 
#@G Karypis
#t2009
#c83
#index313
#id323
#%
#!

#*hierarchical clustering using dynamic modeling clustering discovery process data mining groups set data way maximizes similarity clusters minimizes similarity different clusters advanced algorithms difficulty dealing highly variable clusters model closeness algorithm yields accurate results highly variable clusters existing algorithms use static model clusters use information nature individual clusters furthermore set schemes cure algorithm related schemes information aggregate items clusters set schemes algorithm group method related schemes information closeness clusters defined similarity items clusters considering closeness algorithms select pair clusters key feature accounts closeness identifying similar pair clusters finds clusters data set using phase algorithm phase uses graph partitioning algorithm cluster data items relatively small second phase uses algorithm clusters combining 
#@G Karypis, Eui-Hong Han, Vipin Kumar
#t2009
#c188
#index314
#id324
#%
#!

#*analysis hard soft assignment methods clustering assignment methods heart algorithms unsupervised learning clustering particular known means expectation maximization em algorithms work study different methods assignment including hard used means soft used em known means minimizes distortion data em maximizes likelihood little known systematic differences behavior 
#@M Kearns, Y Mansour
#t2009
#c30
#index315
#id325
#%
#!

#*statistical language modeling text retrieval classification clustering library code useful statistical text analysis language modeling information retrieval programs current distribution includes library document classification document retrieval document clustering 
#@A McCallum
#t2009
#c50
#index316
#id326
#%
#!

#*using unlabeled data improve text classification key difficulty text classification learning algorithms require hand labeled examples learn demonstrates supervised learning algorithms use small number labeled examples inexpensive unlabeled examples create high accuracy text assuming documents created parametric generative model expectation maximization em finds local maximum models classifiers 
#@Nigam, Kamal Paul
#t2009
#c50
#index317
#id327
#%
#!

#*tutorial hidden markov models selected applications speech recognition tutorial provides overview basic theory hidden markov models hmms baum gives practical details methods implementation theory description selected applications theory problems speech recognition results number original sources combined provide single source background required area research author reviews theory discrete markov chains shows concept hidden states observation probabilistic function state used effectively theory illustrated simple examples fundamental problems hmms practical techniques solving problems given various types hmms studied including left right models described 
#@Lawrence R. Rabiner
#t2009
#c181
#index318
#id328
#%
#!

#*deterministic clustering compression classification regression related optimization problems let place neural network perspective particularly area neural networks greatly unique position engineering including probability theory control information theory complexity theory fi neural networks provided 
#@K. Rose
#t2009
#c189
#index319
#id329
#%
#!

#*clustering sequences hidden markov models paper discusses probabilistic model based approach clustering sequences using hidden markov models hmms fi generalization standard approach clustering feature primary novel parameter procedure second difficult problem number clusters data indicate proposed 
#@Padhraic Smyth
#t2009
#c190
#index320
#id330
#%
#!

#*comparison document clustering techniques paper presents results experimental study common document clustering techniques particular compare main approaches document clustering hierarchical clustering means means used standard means algorithm variant means means better quality clustering approach limited quadratic time complexity contrast means variants time complexity linear number documents produce clusters means hierarchical approaches combined best results indicate means technique better standard means approach good better hierarchical approaches tested variety cluster evaluation metrics propose results based analysis clustering algorithms nature document data 
#@Michael Steinbach, G Karypis, Vipin Kumar
#t2009
#c191
#index321
#id331
#%
#!

#*cluster knowledge reuse framework combining partitions paper introduces problem combining multiple set objects single clustering features algorithms determined identify application scenarios knowledge reuse framework cluster cluster problem combinatorial optimization problem terms shared mutual information addition direct maximization approach propose ective efficient techniques obtaining high quality functions similarity measure objects second based partitioning groups clusters meta clusters object determine combined clustering low computational costs techniques quite feasible use function approaches objective function best solution given situation evaluate effectiveness cluster qualitatively different application scenarios original clusters formed based non identical sets features original clustering algorithms non identical sets objects common data set used main purpose combining multiple improve quality robustness solution promising results obtained situations synthetic real data sets 
#@Alexander Strehl, J Ghosh
#t2009
#c128
#index322
#id332
#%
#!

#*relationship based clustering visualization high dimensional data mining real life data mining paper proposes relationship based approach problems dimensionality issue working suitable similarity space instead original high dimensional attribute similarity space satisfy business criteria requiring customer clusters represent comparable amounts apply efficient scalable graph partitioning 
#@Alexander Strehl, J Ghosh
#t2009
#c192
#index323
#id333
#%
#!

#*criterion functions document clustering experiments analysis recent years tremendous growth volume text documents available internet digital libraries news sources wide led increased help users effectively summarize organize information fast high quality document clustering algorithms play role goal shown provide navigation browsing organizing large amounts information small number meaningful clusters greatly retrieval performance cluster driven dimensionality reduction term weighting query expansion increasing importance document clustering range applications led number new novel algorithms different complexity quality trade class algorithms relatively low computational requirements treat clustering optimization process seeks maximize minimize particular clustering criterion function entire clustering solution focus paper evaluate performance different criterion functions problem study involves total seven different criterion functions introduced proposed past evaluation consists comprehensive involving different datasets analysis characteristics various effect clusters produce experimental results set consistently outperform proposed criterion functions lead overall results theoretical analysis criterion function shows relative performance degree operate clusters different degree lead balanced clusters 
#@Ying Zhao, G Karypis
#t2009
#c83
#index324
#id335
#%
#!

#*hmms coupled hmms multi channel eeg classification variety coupled hmms recently proposed hmm better characterize multiple sequences paper introduces novel distance coupled hmm compares performance hmm models multi channel eeg classification problem results approaches examined multivariate hmm low computational complexity outperforms models 
#@Shi Zhong, J Ghosh
#t2009
#c193
#index325
#id336
#%569,570,318,571,572,573,574,575,576,577,578,579,580,581,582,583,584,585,320
#!

#*unified framework clustering applications clustering time sequences model based clustering techniques widely used shown promising results applications involving complex data paper presents unified framework probabilistic model based clustering based bipartite graph view data models differences existing model based clustering algorithms view clusters represented probabilistic models model space conceptually separate data space clustering view conceptually similar em algorithm hierarchical clustering graph based view helps critical important similarity based approaches model based approaches framework suggests useful variations existing clustering algorithms new variations balanced model based clustering hybrid model based clustering discussed empirically evaluated variety data types 
#@Shi Zhong, J Ghosh
#t2009
#c128
#index326
#id337
#%
#!

#*continuous time bayesian networks paper present language finite state continuous time bayesian networks structured stochastic processes continuous time state set local variables values change time dynamics described behavior local variable function directed possibly graph model given point time distribution aspects local variable changes value value takes distributions determined variable current value current values graph variable finite state continuous time markov process transition functions present probabilistic semantics language terms generative model defines sequences events list types queries ask discuss computational difficulties associated exact inference provide algorithm approximate inference takes advantage structure process 
#@Uri Nodelman, Christian R. Shelton, Daphne Koller
#t2009
#c129
#index327
#id338
#%328,329,330,331,332,333,334
#!

#*event history analysis purpose event history analysis explain certain higher risk event using special types methods depending field applied called time models life models transition rate models response time models models models hazard models examples class techniques use terms event history hazard models hazard model regression model risk ing event certain time point predicted set special features hazard models types models make possible deal contain partial information event special feature change observation period including time varying makes possible perform dynamic analysis important types hazard models introduce basic concepts 
#@Paul David Allison
#t2009
#c30
#index328
#id339
#%
#!

#*model reasoning reasoning change requires predicting long having true knowledge agent constrained believe simply way agent infer paper model reasoning accounts knowledge concerning effect relationships knowledge concerning function time model natural encoding form network representation probabilistic models consider computational properties model recent advances computing models network representation finally discuss probabilistic model addresses certain classical problems temporal reasoning problems 
#@Thomas Dean, Keiji Kanazawa
#t2009
#c194
#index329
#id340
#%
#!

#*recursive resolution uncertainty derive implications risk setting rate hazard rate depend market value itself market values entity determined simultaneously key technique use recursive stochastic equations characterize market value manner resolution uncertainty particular conditions value respect information 
#@Darrell Duffie, Mark Schroder, Costis Skiadas
#t2009
#c195
#index330
#id341
#%
#!

#*probabilistic temporal reasoning change paper presents probabilistic model reasoning state changes time target domain class prediction problems diagnosis allow arbitrary testing domains typically time gather information state consider alternative 
#@Steve Hanks, David Madigan, Jonathan Gavrin
#t2009
#c129
#index331
#id342
#%
#!

#*processes credit framework presented modeling credit allows market risk factors credit risk framework reduces technical issues modeling credit risk issues modeling term structure rates shown model allow stochastic transition rating categories generalization handle explicitly linked ratings shown obtain term structure model different rating categories simultaneously obtain affine like structure implementation given simple factor model affine structure gives closed form solutions 
#@David Lando
#t2009
#c196
#index332
#id343
#%
#!

#*probabilistic reasoning intelligent systems probabilistic reasoning intelligent systems complete account theoretical foundations computational methods reasoning uncertainty author provides probability language reasoning partial belief offers unifying perspective approaches uncertainty systems logic author syntactic semantic approaches uncertainty offers techniques based belief networks provide mechanism making semantics based systems specifically network propagation techniques serve mechanism combining theoretical probability theory modern reasoning systems technology inputs conceptually meaningful parallel distributed computation application areas include diagnosis image interpretation multi sensor decision support systems plan recognition planning speech recognition short task requiring clues information probabilistic reasoning intelligent systems special researchers decision theory statistics logic management sciences areas knowledge based systems operations research engineering statistics theoretical computational tools immediate practical use book used excellent text level operations research applied probability 
#@Judea Pearl
#t2009
#c30
#index333
#id344
#%
#!

#*probability propagation paper simple account local computation joint probability distribution given sets variables involved factors form local computation conditional probability believe emphasis essential local computation factorization essential factorization terms probabilities account given required probabilities alternative measures probability belief functions 
#@Glenn Shafer, Prakash P. Shenoy
#t2009
#c197
#index334
#id345
#%
#!

#*correlated topic models topic models latent dirichlet allocation lda useful tools statistical analysis document collections discrete data lda model assumes words document arise mixture topics distribution vocabulary limitation lda model topic correlation example document likely limitation use dirichlet distribution model variability topic paper develop correlated topic model topic exhibit correlation normal distribution derive mean field variational inference algorithm approximate posterior inference model fact normal conjugate multinomial gives better lda collection articles journal science furthermore provides natural way exploring unstructured data sets 
#@David M. Blei, J Lafferty
#t2009
#c190
#index335
#id346
#%336,176,337,338,339,340,341,342,343
#!

#*variational inference engine bayesian networks recent years variational methods popular tool approximate inference learning wide variety probabilistic models new application currently necessary derive variational update equations implement application specific code steps time consuming error paper general purpose inference engine called variational inference bayesian networks allows wide variety probabilistic models implemented solved coding new models specified simple graphical interface drawing automatically variational equations illustrate power using examples bayesian mixture modelling 
#@Christopher M. Bishop, David J. Spiegelhalter, J. M. Winn
#t2009
#c190
#index336
#id347
#%203,344,345,346,190,347,348
#!

#*mixtures markov chains distributed modelling dynamic user provide compact generative representation sequential activity number individuals group tion individual global models paper proposes linear time distributed model state sequences representing individual user activity making assumption heterogeneous user relatively small number simple common patterns user results empirical study different sources user indicates modelling approach provides cient representation scheme improved prediction performance providing low complexity intuitively representations 
#@Mark Girolami, Ata Kaban
#t2009
#c190
#index337
#id348
#%349,350,351,352,353,354
#!

#*integrating topics statistical approaches language learning typically focus short range syntactic dependencies long range semantic dependencies words present generative model uses kinds dependencies simultaneously finding syntactic classes semantic topics despite having knowledge semantics statistical model competitive tasks like speech tagging document classification models use short long range dependencies respectively 
#@Thomas L. Griffiths, Mark Steyvers, David M. Blei, Joshua B. Tenenbaum
#t2009
#c190
#index338
#id349
#%355,356,357,358,176,359,360,361,362,363
#!

#*collaborative filtering machine learning perspective collaborative filtering proposed framework filtering information based preferences users different ways comprehensive study rating based pure non sequential collaborative filtering analyze existing methods task rating prediction machine learning perspective existing methods proposed task simple applications standard machine learning methods classification regression clustering dimensionality reduction density estimation introduce new prediction methods classes introduce new experimental procedure testing forms generalization used previously implement total prediction methods conduct large scale prediction accuracy experiments interesting new results relative performance methods 
#@Benjamin Marlin, Benjamin Marlin
#t2009
#c50
#index339
#id350
#%
#!

#*inference population structure using data model based clustering method using data infer population structure assign individuals populations assume model populations unknown characterized set frequencies individuals sample populations populations indicate model does assume particular process applied commonly used genetic provided closely linked applications method include presence population structure assigning individuals populations studying hybrid identifying individuals method produce highly accurate using numbers seven example using data software used article available www home html 
#@JK Pritchard, M. Stephens, P. Donnelly
#t2009
#c19
#index340
#id351
#%
#!

#*discovering object categories image collections given set images containing multiple object categories seek discover categories image locations achieve using generative models statistical text literature probabilistic latent semantic analysis plsa latent dirichlet allocation lda text analysis used discover topics corpus using bag words document representation discover topics object categories image containing instances categories mixture topics models applied images using visual word formed vector like region descriptors investigate set increasingly scenarios starting image sets containing object categories sets containing multiple categories including background object categories sample intra class scale variation categories approximate spatial layout demonstrate unseen images images containing multiple objects performance proposed unsupervised method compared semi supervised approach work project university intelligence 
#@Josef Sivic, Bryan C. Russell, Alexei A. Efros, Andrew Zisserman, William T. Freeman
#t2009
#c50
#index341
#id352
#%364,365,176,366,367,368,360,199,349,369,370,371,372,373,374,375,376,377,378
#!

#*variational principle graphical models graphical models graph theory probability theory powerful multivariate statistical modeling statistical signal processing related fields communication theory control theory statistical models long formulated terms graphs algorithms computing basic statistical quantities probabilities expressed terms operating 
#@H Attias
#t2009
#c30
#index342
#id353
#%
#!

#*generalized mean field algorithm variational inference exponential present class generalized mean field algorithms approximate inference exponential family graphical models generalized belief propagation cluster variational methods based 
#@Eric P. Xing, Michael I. Jordan, Stuart J. Russell
#t2009
#c129
#index343
#id354
#%
#!

#*new view em algorithm incremental variants present new view em algorithm maximum likelihood situations unobserved view algorithm seen joint function model distribution unobserved perspective incremental variant algorithm unobserved variables shown 
#@Radford Neal, Geoffrey E. Hinton
#t2009
#c198
#index344
#id355
#%
#!

#*winbugs bayesian modelling framework concepts structure winbugs fully framework constructing bayesian probability models models specified language using graphical interface called winbugs processes model object oriented representation model software offers user interface based model using markov chain monte carlo techniques paper discuss various modern computing concepts object orientation run time linking feature software design discuss framework extended possible specific applications form interface winbugs users specialized requirements possible interface winbugs lower level incorporating new object types used winbugs knowledge implemented types extension require access winbugs source code 
#@David J. Lunn, Andrew Thomas, Nicky Best, David Spiegelhalter
#t2009
#c199
#index345
#id356
#%
#!

#*propagation algorithms variational bayesian learning variational approximations widespread tool learning graphical modelsfi provide variational updates general family exponential graphical modelsfi tree algorithms used inference step variational bayesian applying bayesian analysis linear gaussian state space obtain learning procedure 
#@Zoubin Ghahramani, Matthew J. Beal
#t2009
#c190
#index346
#id357
#%
#!

#*variational principal components central issues use component analysis pca data choosing appropriate components problem addressed formulation pca probabilistic latent variable model feature approach dimensionality latent space number principal com determined automatically bayesian inference procedure com non bayesian models required intractable based local gaussian tion posterior distribution employed paper develop alternative formulation bayesian pca based representation posterior approach computationally cient unlike approximation schemes maximizes lower bound log probability observed data 
#@Christopher M. Bishop
#t2009
#c200
#index347
#id358
#%
#!

#*structured variational distributions variational methods approximate solution models machine learning vision information retrieval unfortunately new necessary derive variational update equations particular probabilistic model used implement equations steps 
#@Christopher M. Bishop, J. M. Winn
#t2009
#c201
#index348
#id359
#%
#!

#*unsupervised learning probabilistic latent semantic analysis paper presents novel statistical method factor analysis binary data closely related technique known latent semantic analysis contrast method linear algebra performs singular value decomposition occurrence tables uses generative latent class model perform probabilistic mixture decomposition principled approach foundation statistical inference propose use version expectation maximization algorithm model shown excellent performance practice probabilistic latent semantic analysis applications information retrieval natural language processing machine learning text related areas paper presents perplexity results different types text linguistic data collections discusses automated document indexing experiments indicate substantial consistent probabilistic method standard latent semantic analysis 
#@Thomas Hofmann
#t2009
#c23
#index349
#id360
#%
#!

#*maximum likelihood estimation dirichlet distributions dirichlet distributions commonly used data paper introduce distribution discuss useful compare implementations different methods estimating parameters observed data 
#@G Ronning
#t2009
#c202
#index350
#id361
#%668,176
#!

#*maximum likelihood estimation model discrete time model special mixture independent markov chains widely used modeling dynamics social processes problem maximum likelihood estimation parameters data sample independent process considered literature present maximum likelihood procedure estimation parameters model develop recursive method computation maximum likelihood estimators simple implement obtained maximum likelihood estimators strongly consistent estimators parameters model previously proposed literature special cases maximum likelihood estimator derived article maximum likelihood estimator special conditions explain interconnection existing estimators present numerical comparison estimators finally illustrate application maximum likelihood estimators testing hypothesis markov chain describes data hypothesis model describes data 
#@Halina Frydman
#t2009
#c133
#index351
#id362
#%
#!

#*algorithms non negative matrix factorization non negative matrix factorization previously shown useful decomposition multivariate data different multi algorithms analyzed factor used update rules algorithm shown minimize conventional error minimizes generalized convergence algorithms proven using tion used convergence expectation maximization algorithm algorithms gradient factor chosen ensure convergence 
#@Daniel D. Lee, H. Sebastian Seung
#t2009
#c190
#index352
#id363
#%
#!

#*expectation generative aspect model generative aspect model extension multinomial model text allows vary documentsfi 
#@Thomas P. Minka, J Lafferty
#t2009
#c129
#index353
#id364
#%
#!

#*learning making model setting tool used make predictions model seen makes quantities inferred effect bayesian framework knowledge contained conditional modelsfi use bayes theorem 
#@Robi Polikar
#t2009
#c50
#index354
#id365
#%
#!

#*language different neural different sensitive theoretical considerations studies provided support semantic grammatical functions language domain evidence concerning hypothesis comparing normal event related brain words provide primarily semantic information open class grammatical information closed class comparing effects early language experience subjects open closed class words normal different word types qualitatively different different word classes language processing addition indices semantic processing identical subjects linked grammatical processes different subjects results suggest neural systems different different aspects language generally results provide support semantic grammatical functions 
#@HJ Neville, DL Mills, DS Lawson
#t2009
#c203
#index355
#id366
#%
#!

#*distributional information powerful syntactic categories idea distributional information play significant role syntactic category demonstrate empirically information provides powerful syntactic category membership variety simple mechanisms present range results using large corpus directed speech explore implications results considerable information concerning syntactic categories obtained distributional information sources information potential identification syntactic classes 
#@Martin Redington, Nick Chater, Steven Finch
#t2009
#c204
#index356
#id367
#%
#!

#*solution problem latent semantic analysis theory induction representation knowledge people know little information problem takes forms learning vocabulary text especially convenient case research new general theory acquired similarity knowledge representation latent semantic analysis lsa presented used successfully learning phenomena global knowledge local occurrence data large representative text lsa acquired knowledge vocabulary english comparable rate lsa uses prior linguistic similarity knowledge based general mathematical learning method achieves powerful inductive effects extracting right number dimensions represent objects contexts relations phenomena problems 
#@TK Landauer, S. Dumais
#t2009
#c205
#index357
#id368
#%
#!

#*foundations statistical natural language processing statistical approaches processing natural language text recent years text comprehensive introduction statistical natural language processing nlp appear book contains theory algorithms needed building nlp tools provides broad coverage mathematical linguistic foundations detailed discussion statistical methods allowing students researchers construct implementations book covers finding word sense disambiguation probabilistic parsing information retrieval applications 
#@Paul B. Kantor
#t2009
#c206
#index358
#id369
#%
#!

#*better integration semantic predictors statistical language modeling introduce number techniques designed help integrate semantic knowledge gram language models automatic speech recognition techniques allow integrate latent semantic analysis lsa word similarity algorithm based word occurrence information gram models lsa good predicting content words text bad predictor frequent words low dynamic range combined grams dynamic range applying word confidence metric using geometric linear grams produces robust language model lower perplexity journal baseline gram model 
#@Noah Coccaro, Daniel Jurafsky
#t2009
#c207
#index359
#id370
#%606,65,607,608,609,610,611
#!

#*finding scientific topics step identifying content document determining topics document addresses generative model documents introduced document generated choosing distribution topics choosing word document topic selected according distribution present markov chain monte carlo algorithm inference model use algorithm analyze using bayesian model selection establish number topics extracted topics capture meaningful structure data consistent class provided authors articles outline applications analysis including identifying topics examining temporal dynamics tagging illustrate semantic content 
#@Thomas L. Griffiths, Mark Steyvers
#t2009
#c208
#index360
#id371
#%
#!

#*markov chain monte carlo practice markov chain monte carlo methods make use flexible bayesian models computationally recent years applications described applied new questions concerns effort needed design use chain confidence answers produces does use affect model building process joint statistical users discussed issues various trade article discussion purpose users users topics include building simulation results methods estimating standard errors identification good algorithms exist state software development 
#@WR Gilks, DJ Spiegelhalter
#t2009
#c209
#index361
#id372
#%
#!

#*bayes factors paper book theory probability developed evidence scientific theory number called bayes factor posterior hypothesis discussion testing context values attention given factor practical tool 
#@Robert E. Kass, Adrian E. Raftery
#t2009
#c133
#index362
#id373
#%
#!

#*comparing partitions problem comparing different partitions finite set objects clustering literature begin known measure partition discuss issue correcting index recent strategy developed adopted based assumption general problem comparing partitions assessing proximity matrices using simple cross product measure generated corresponding partitions using various scoring rules special cases include traditionally statistics ones weight certain object pairs finally propose measure based comparison object having advantage probabilistic interpretation addition assuming constant value reasonable hypothesis bounded 
#@Lawrence Hubert, Phipps Arabie
#t2009
#c210
#index363
#id374
#%
#!

#*matching words present new rich approach multi sets focusing specific case images learning joint image regions images annotation particular image regions region fi annotation organize access large collections region 
#@Kobus Barnard, Pinar Duygulu, David A. Forsyth, Nando de Freitas, David M. Blei, Michael I. Jordan
#t2009
#c128
#index364
#id375
#%
#!

#*minimum complexity density estimation authors introduce index bound rate convergence minimum complexity density estimators information theoretic redundancy corresponding total description length results index demonstrate statistical effectiveness minimum description length principle method inference minimum complexity estimator true density nearly fast estimator based prior knowledge true basic properties minimum complexity estimators discussed regression classification problems examined minimum description length framework considered 
#@Andrew R. Barron, TM Cover
#t2009
#c211
#index365
#id376
#%
#!

#*visual present novel method generic visual categorization problem identifying object content natural images variations object class bag method based vector quantization affine invariant descriptors image propose compare alternative implementations using different classifiers bayes svm main advantages method simple computationally efficient invariant present results simultaneously classifying seven semantic visual categories results clearly demonstrate method robust background produces good categorization accuracy exploiting geometric information 
#@Gabriella Csurka, Christopher R. Dance, Lixin Fan, Jutta Willamowski, Cdric Bray
#t2009
#c212
#index366
#id377
#%
#!

#*object recognition machine learning lexicon fixed image vocabulary model object recognition machine 
#@Pinar Duygulu, Kobus Barnard, Joao F. G. de Freitas, David A. Forsyth
#t2009
#c213
#index367
#id378
#%
#!

#*learning generative visual models training examples incremental bayesian approach tested object categories current computational approaches learning visual object categories require thousands training images slow learn incremental manner incorporate prior information learning process addition algorithm presented literature tested object categories present method learning object categories just training images uses prior information principled way test dataset images objects widely categories proposed method based making use prior information object categories previously generative probabilistic model used represents shape features object parameters model bayesian manner incremental algorithm compared experimentally earlier batch bayesian algorithm based maximum likelihood incremental batch versions comparable classification performance small training sets incremental learning significantly faster making real time learning feasible bayesian methods outperform maximum likelihood small training sets 
#@Fei-Fei Li, Robert Fergus, Pietro Perona
#t2009
#c214
#index368
#id379
#%
#!

#*robust wide baseline regions wide baseline problem problem pair images taken different studied new set image elements called regions introduced properties set closed transformation image age efficient near linear complexity practically fast algorithm near rate presented subset regions regions new robust similarity measure proposed robustness multiple regions regions obtained invariant regions significantly larger used establish high utility multiple regions robust metric demonstrated wide baseline experiments image pairs significant change scale conditions scale change present test good estimates geometry average distance points line obtained 
#@Jiri Matas, Ondrej Chum, Martin Urban, Tomas Pajdla
#t2009
#c215
#index369
#id380
#%
#!

#*weak hypotheses boosting generic object detection recognition fi paper stage new learning object detection propose boosting underlying learning allows use diverse sets visual features learning process com framework boosting weak hypotheses choose features relevant tion final advantage weak hypotheses search weak hypotheses space explicit calculation available hypotheses reducing computation related work used learning algorithm weak hypotheses empirical evaluation use types local descriptors basic ones consisting set intensity high level descriptors fi descriptors local inter point weak hypotheses selects local type local searches similarity di work boosting object recognition simple hypotheses complex classifiers relatively simple images objects approach yields results comparable state art fi obtain good results complex images objects arbitrary scales results indicate flexible approach allows features segmented spatial relationships leads significant step generic object 
#@Andreas Opelt, Michael Fussenegger, Axel Pinz, Peter Auer
#t2009
#c216
#index370
#id381
#%
#!

#*statistical method object detection applied paper statistical method object represent statistics object appear non object using product represents joint statistics subset coefficients position approach use histograms representing wide variety visual using method algorithm detect human algorithm detect wide range 
#@Henry Schneiderman, Takeo Kanade
#t2009
#c217
#index371
#id382
#%
#!

#*video google text retrieval approach object matching videos approach object retrieval searches occurrences user object video object represented set invariant region descriptors recognition successfully despite changes partial temporal video used track regions order regions reduce effects noise descriptors text retrieval implementation matches descriptors pre computed using vector quantization inverted file systems document rankings used result retrieval immediate ranked list key manner google method illustrated matching length feature 
#@Josef Sivic, Andrew Zisserman
#t2009
#c218
#index372
#id383
#%
#!

#*automated image retrieval using color texture growing digital images videos increasing need filters based visual content unfortunately effective tools searching retrieving images videos large images videos digital form addition text based indexes built human digital image video libraries require new algorithms automated extraction indexing relevant image features paper investigate automated content extraction database searching based color texture features features important fundamental characteristics content images work general application databases images videos variety domains present new algorithms automated extraction color texture information use binary set representations color texture respectively special emphasis processes feature extraction indexing demonstrate binary feature sets texture color provide excellent performance query response time maintaining highly effective accuracy spatial extraction compressed data representations present color texture set extraction indexing techniques contrast approaches examine texture color searching databases color images respectively finally examine relationship color texture application image video retrieval explore combining color texture maps images obtain unified feature maps characterize image regions color texture particular indicate combining better capture index color patterns characterize real world objects images videos finally examine nature performance image video database queries combine color texture 
#@J.R. Smith, S.F. Chang
#t2009
#c219
#index373
#id384
#%
#!

#*improved video content indexing multiple latent semantic analysis low level features build efficient content based retrieval systems users interested longer retrieving similar content retrieval systems documents similar semantic content gap low level features semantic content challenging task necessary future retrieval systems latent semantic analysis lsa successfully introduced efficiently index text documents detecting words successfully proposed adaptation lsa model video content object retrieval semantic content estimation following idea present new model multiple lsa better represent video content experimental make comparison lsa lsa problems object retrieval semantic content estimation 
#@Fabrice Souvannavong, Bernard Merialdo, Benoit Huet
#t2009
#c220
#index374
#id385
#%
#!

#*contextual object detection general context rich source information object identity location fact structure real world strong apply single introduce simple framework relationship context object properties based correlation low level features entire objects 
#@Antonio B. Torralba
#t2009
#c221
#index375
#id386
#%
#!

#*rapid object detection using simple features paper describes machine learning approach visual object detection processing images extremely rapidly achieving high detection work key introduction new image representation called image allows features used computed second learning algorithm based selects small number critical visual features 
#@Paul A. Viola, Michael J. Jones
#t2009
#c222
#index376
#id387
#%
#!

#*unsupervised learning models recognition fi present method learn object class models unlabeled purpose visual object particular type model objects represented flexible parts features fi variability class joint probability density function pdf shape output stage method automatically 
#@Markus Weber, Max Welling, Pietro Perona
#t2009
#c223
#index377
#id388
#%
#!

#*hidden semantic concept discovery region based image retrieval work addresses content based image retrieval focusing developing hidden semantic concept discovery methodology address effective semantics image retrieval approach image database segmented region associated color texture shape features exploiting statistical information image vector quantization method uniform sparse region based representation achieved representation probabilistic model based statistical hidden class assumptions image database obtained expectation maximization em technique applied analyze semantic concepts hidden database retrieval algorithm designed support probabilistic model semantic similarity measured integrating posterior probabilities transformed query image constructed negative example discovered semantic concepts proposed approach statistical foundation experimental evaluations database general images demonstrate effectiveness 
#@Ruofei Zhang, Zhongfei (Mark) Zhang
#t2009
#c224
#index378
#id389
#%
#!

#*novel review ranking systems rapid growth internet users ability content created active electronic communities provide wealth product information consumers naturally reviews order decide buy product high volume reviews typically published single product makes individuals best reviews understand true underlying quality product based reviews manufacturer product needs identify reviews customer base examine content reviews paper propose ranking mechanisms ranking product reviews consumer oriented ranking mechanism ranks reviews according expected ranking mechanism ranks reviews according expected sales ranking mechanism combines analysis text mining techniques general subjectivity analysis particular subjectivity analysis useful clues review impact sales results implications market design online opinion 
#@Anindya Ghose, Panagiotis G. Ipeirotis
#t2009
#c225
#index379
#id390
#%380,381,382,383,384,179,385,139,386,142,189,387,388
#!

#*power product features mining consumer reviews increasing internet dramatically way consumers product reviews valuable source information customers read reviews decide buy product based information provided paper use techniques reviews segments evaluate individual characteristics product image quality life digital major contribution paper methods literature specifically regression concept estimate weight customers place individual product feature implicit evaluation score customers assign feature evaluations given product goal develop novel hybrid technique combining text mining models consumer product reviews elements product feature evaluation spaces quantitative impact consumer reviews product linear functional product space demonstrate use approximation functional reduce number model parameters providing good experimental results evaluate technique using data set amazon com consisting sales data related consumer reviews period products experimental evaluation shows extract business intelligence data better understand customer preferences actions textual portion reviews improve product sales prediction compared baseline technique simply numeric data 
#@Nikolay Archak, Anindya Ghose, Panagiotis Ipeirotis
#t2009
#c3
#index380
#id391
#%
#!

#*effect word sales online book reviews creation online consumer communities provide product reviews important expensive component internet retail strategies paper characterize behavior popular internet sites examine effect consumer reviews sales use available data leading online amazon com com construct measures sales individual books gather extensive consumer review data sites characterize behavior sites distribution number ratings length ratings ratings different subject categories second measure effect individual reviews relative books sites argue methodology comparing sales reviews given book internet allows improve existing literature better capturing relationship word reviews sales able difference factors affect sales word book quality examine incremental sales effects having reviews particular book versus having reviews differential sales effects positive negative reviews large database books allows control important factors differences sites times 
#@JA Chevalier, D Mayzlin
#t2009
#c226
#index381
#id392
#%
#!

#*yahoo amazon sentiment extraction small web extracting sentiment text hard semantic problem develop methodology extracting small sentiment stock message algorithm different classifier algorithms coupled voting scheme accuracy levels similar widely used bayes classifiers false lower sentiment accuracy higher time series cross aggregation message information improves quality sentiment index particularly presence empirical applications evidence relationship stock postings related stock index levels algorithms used assess impact opinion management party news changes 
#@SR Das, MY Chen
#t2009
#c227
#index382
#id393
#%
#!

#*multi level impact social economic transactions electronic important uses internet today economic social interaction source information paper explore come form social communities built information electronic drawing social identity theory suggest relationship online consumer reviews internet product sales social identity processes using unique dataset based compiled ratings characteristics given set products location based behavior amazon provide evidence community level linking identity behavior online community subsequent product sales addition particular location subsequent product sales higher region review level analysis subsequent reviews identity norms set previous reviews identity community norms identity furthermore results suggest identity implications economic form product sales implications research word electronic communities discussed 
#@Forman Chris, Ghose Anindya, Wiesenfeld Batia
#t2009
#c228
#index383
#id394
#%
#!

#*opinion mining using case study systems polarity strength opinions important research topic significant attention years work measure strength polarity opinion consider economic context opinion evaluated instead using human linguistic resources rely fact text line systems behavior humans effect observed using easy measure economic variables product logic infer semantic orientation strength opinion changes associated economic variable effect use identify economic value text assign value opinion phrase measuring sentiment effectively need manual labeling argue opinions using objective evaluation opinions make discussion presenting results amazon com user feedback power measuring power infer polarity strength underlying feedback postings 
#@Anindya Ghose, Panagiotis G. Ipeirotis, Arun Sundararajan
#t2009
#c104
#index384
#id395
#%
#!

#*mining summarizing customer reviews selling products web ask customers review products associated services commerce popular number customer reviews product grows rapidly popular product number reviews hundreds thousands makes difficult potential customer read make informed decision purchase product makes difficult manufacturer product track customer opinions manufacturer additional difficulties sites product manufacturer produces kinds products research aim summarize customer reviews product summarization task different traditional text summarization features product customers expressed opinions opinions positive negative summarize reviews selecting subset original sentences reviews capture main points text summarization task performed steps mining product features customers identifying opinion sentences review opinion sentence positive negative summarizing results paper proposes novel techniques perform tasks experimental results using reviews number products online demonstrate effectiveness techniques 
#@Minqing Hu, Bing Liu
#t2009
#c3
#index385
#id396
#%389,390,391,392,393,394,395,396,397,398,399,400,401,402,403,404,405,358,406,145,407,408,409
#!

#*market agents better informed value information real estate transactions agents better informed exploit advantage real estate agents know market typical example real estate agents small share incremental profit higher value quickly test predictions comparing home sales real estate agents home instances real estate agent home case agent case agent wants best consistent theory real estate agents market longer wide range characteristics situations agent advantage larger lead 
#@Steven D. Levitt, Chad Syverson
#t2009
#c229
#index386
#id397
#%
#!

#*sentiment classication using machine learning techniques consider problem classifying documents topic overall sentiment determining review positive negative using movie reviews data standard machine learning techniques outperform human produced machine learning methods employed naive bayes maximum entropy classification support vector machines perform sentiment classification traditional topic based categorization conclude examining factors make sentiment classification problem challenging 
#@Bo Pang, Lillian Lee, Shivakumar Vaithyanathan
#t2009
#c107
#index387
#id398
#%
#!

#*thumbs thumbs semantic orientation applied unsupervised classication reviews paper presents simple unsupervised learning algorithm classifying reviews recommended thumbs recommended thumbs classification review predicted average semantic orientation phrases review contain adjectives phrase positive semantic orientation good associations negative semantic orientation bad associations paper semantic orientation phrase mutual information given phrase word excellent mutual information given phrase word poor review classified recommended average semantic orientation phrases positive algorithm achieves average accuracy evaluated reviews different domains reviews movies accuracy reviews movie reviews 
#@Peter Turney
#t2009
#c230
#index388
#id399
#%
#!

#*recognizing subjectivity case study manual tagging paper case study sentence level categorization tagging developed used classify journal subjective objective agreement analyzed based analysis given final classification provide empirical support classifications correlations data subjective category basic semantic class 
#@Rebecca Bruce, Janyce Wiebe
#t2009
#c161
#index389
#id400
#%
#!

#*combining low level summary representations opinions multi perspective question answering recent research fact based question answering work aims extend question answering research different direction handle multi perspective question answering tasks question answering tasks require ability organize opinions text particular paper proposes approach multi perspective question answering views task opinion oriented information extraction 
#@Claire Cardie, Janyce Wiebe, Theresa Wilson, Diane J. Litman
#t2009
#c231
#index390
#id401
#%
#!

#*word association norms mutual information term word association used particular sense literature generally subjects normal word follows highly associated word extend term provide basis statistical description variety interesting linguistic phenomena ranging semantic relations type content word content word syntactic occurrence constraints content word function word paper propose objective measure based information theoretic notion mutual information estimating word association norms corpora standard method obtaining word association norms testing subjects words costly proposed measure association ratio estimates word association norms directly corpora making possible estimate norms thousands words 
#@Kenneth Ward Church, Patrick Hanks
#t2009
#c104
#index391
#id402
#%
#!

#*study implementation combined techniques automatic extraction paper presents original method implementation extract corpora combining linguistic filters statistical methods starting linguistic study terms domain designed number filters enable obtain selection sequences considered terms various statistical scores applied selection results evaluated method applied english 
#@Batrice Daille
#t2009
#c232
#index392
#id403
#%
#!

#*mining opinion extraction semantic classification product reviews web contains wealth product reviews task opinion mining tool process set search results given item generating list product attributes quality features opinions poor mixed good begin identifying unique properties problem develop method automatically distinguishing positive negative reviews classifier information retrieval techniques feature extraction scoring results various metrics heuristics vary depending testing situation best methods work better traditional machine learning operating individual sentences collected web searches performance limited noise context complete web based tool simple method grouping sentences attributes results qualitatively quite useful 
#@K Dave, S Lawrence, DM Pennock
#t2009
#c38
#index393
#id404
#%
#!

#*wordnet electronic lexical database wordnet important widely used lexical resource natural language processing systems wordnet electronic lexical database discusses design wordnet theoretical historical provides date description lexical database presents set applications wordnet book contains introduction seven chapters sciences university wordnet produced chapters 
#@Christiane Fellbaum
#t2009
#c30
#index394
#id405
#%
#!

#*genre classification domain transfer information filtering world wide web information volume makes identify useful documents identify document genre important factor retrieving useful documents focus novel document genre dimension subjectivity investigate approaches automatically classifying documents genre traditional bag words techniques speech statistics hand linguistic features particularly interested domain transfer learned classifiers training corpus new document corpus experiments demonstrate speech approach better traditional bag words techniques particularly domain transfer conditions 
#@Aidan Finn, Nicholas Kushmerick, Barry Smyth
#t2009
#c34
#index395
#id406
#%
#!

#*summarizing text documents sentence selection evaluation metrics human quality text summarization systems difficult design difficult evaluate documents dimensions length style lexical usage certain help suggest selection sentences summary paper presents analysis news article summaries generated sentence selection sentences ranked potential summary using weighted 
#@Jade Goldstein, Mark Kantrowitz, Vibhu O. Mittal, Jaime G. Carbonell
#t2009
#c35
#index396
#id407
#%
#!

#*effects orientation sentence subjectivity subjectivity sentence level feature important implications processing information information study adjectives semantically oriented adjectives ad simple subjectivity establish arc strong predictors subjectivity novel statistically combines presented automatic assigning orientation labels 
#@Vasileios Hatzivassiloglou, Janyce Wiebe
#t2009
#c233
#index397
#id408
#%
#!

#*direction based text interpretation information access refinement text based intelligent provide depth information contents corpus does standard information retrieval time complexity resource consuming behavior detailed text instead focusing discovering documents topic user approach introduced based criterion agent event method described sentence model semantic interpretation needed order determine sentence respect model interpretation method designed integrated component hybrid information access 
#@Marti A. Hearst
#t2009
#c234
#index398
#id409
#%
#!

#*mining opinion features customer reviews common practice selling products web ask customers review products associated services commerce popular number customer reviews product grows rapidly popular product number reviews hundreds makes difficult potential customer read order make decision buy product project aim summarize customer reviews product summarization task different traditional text summarization interested specific features product customers opinions opinions positive negative summarize reviews selecting subset original sentences reviews capture main points text summarization paper focus mining opinion product features number techniques presented features experimental results techniques highly effective 
#@Minqing Hu, Bing Liu
#t2009
#c103
#index399
#id410
#%
#!

#*fuzzy typing document management prototype demonstrates novel method document analysis management based combination techniques nlp fuzzy logic central technique use nlp semantic typing refer approach fuzzy typing document management 
#@Alison Huettner, Pero Subasic
#t2009
#c19
#index400
#id411
#%
#!

#*term extraction automatic indexing chapter presents new domain research development natural language processing nlp concerned representation recognition terms terms scientific technical documents identification crucial issue application dealing analysis understanding generation documents particular growing specialized available line 
#@Ruslan Mitkov
#t2009
#c30
#index401
#id412
#%
#!

#*recognizing text genres simple metrics using analysis simple method texts pre determined text genre categories using statistical standard technique analysis demonstrated application corpus analysis makes possible use large number parameters specific certain corpus information stream combine small number functions parameters weighted basis useful text genres application information retrieval discussed 
#@Jussi Karlgren, Douglas R. Cutting
#t2009
#c233
#index402
#id413
#%
#!

#*automatic detection text genre text databases available users larger heterogeneous genre increasingly important computational linguistics topical structural principles classification propose theory genres facets various surface argue genre detection based surface successful detection based structural properties 
#@Brett Kessler, Geoffrey Nunberg, Hinrich Schutze
#t2009
#c104
#index403
#id414
#%
#!

#*integrating classification association rule mining concept efficient tool data analysis paper classification association rule mining unified concept framework present fast algorithm extract association classification rules concept 
#@Bing Liu, Wynne Hsu, Yiming Ma
#t2009
#c3
#index404
#id415
#%
#!

#*multi document summarization graph search matching new method summarizing similarities differences pair related documents using graph representation text concepts words phrases proper names document represented nodes graph edges corresponding semantic relations items given perspective terms pair documents algorithm uses activation technique discover document nodes semantically related topic graphs document yield graph corresponding similarities differences pair natural language evaluation techniques carried 
#@Inderjeet Mani, Eric Bloedorn
#t2009
#c175
#index405
#id416
#%
#!

#*introduction wordnet line lexical database wordnet line lexical reference design current human lexical memory english nouns adjectives sets representing underlying lexical concept different relations link sets 
#@George A. Miller, Richard Beckwith, Christiane Fellbaum, Derek Gross, Katherine J. Miller
#t2009
#c235
#index406
#id417
#%
#!

#*automatic text decomposition using text segments text themes widespread use text information retrieval retrieval techniques increasingly popular larger texts important text retrieval task improving retrieval effectiveness evidence use words local contexts useful language improving retrieval output main text decomposition strategies introduced study including decomposition text segments semantic decomposition text themes interaction text segments text themes used characterize text structure information retrieval text text summarization 
#@Gerard Salton, Amit Singhal, Chris Buckley, Mandar Mitra
#t2009
#c236
#index407
#id418
#%
#!

#*analysis based resolution methods paper presents resolution algo constraints rules theory et al alternative et al algorithm advantages new model left right algorithm incremental processing low computational overhead algorithm compared resolution methods based algorithm list approach 
#@Joel R. Tetreault
#t2009
#c104
#index408
#id419
#%
#!

#*learning subjective adjectives corpora subjectivity tagging distinguishing sentences used present opinions evaluations sentences used present information numerous applications subjectivity tagging relevant including information extraction information retrieval paper identifies strong clues subjectivity using results method clustering words according distributional similarity lin small detailed manual annotation 
#@Janyce Wiebe
#t2009
#c175
#index409
#id420
#%
#!

#*latent semantic indexing lsi trec report paper reports recent semantic indexing lsi retrieval method fi lsi uses reduced dimension vector represent words documentsfi representation terms automatically used improve used lsi trec routing routing tasks lsi space using training documentsfi 
#@S. Dumais
#t2009
#c52
#index410
#id421
#%
#!

#*topic based language models using em paper propose novel statistical language model capture topic related long range dependencies topics modeled latent variable framework derive em algorithm perform topic factor decomposition based segmented training corpus topic model combined standard language model used line word prediction perplexity results indicate improvement previously proposed topic models unfortunately lower word error 
#@Daniel Gildea, Thomas Hofmann
#t2009
#c237
#index411
#id422
#%
#!

#*latent class models collaborative filtering paper presents statistical approach collaborative use latent class models predicting choices preferences based observed preference behavior models discussed compared aspect model latent space model models preferences combination preference factors clustering model simultaneously partitions objects clusters present em algo different variants aspect model derive approximate em algorithm based variational principle model different experimentally investigated large movie data set 
#@Thomas Hofmann, Jan Puzicha
#t2009
#c238
#index412
#id423
#%
#!

#*unsupervised learning data data refers domain finite sets objects pairs element set includes event occurrences data single preference data special cases data arises naturally applications ranging computational linguistics information retrieval preference analysis vision paper present systematic domain independent framework unsupervised learning data statistical mixture models approach covers di models hierarchical latent class structures probabilistic modeling structure discovery mixture models provide probability distributions good generalization sparse data structural information data grouping structure propose version standard expectation maximization algorithm model empirically evaluated variety data sets different domains 
#@Thomas Hofmann, Jan Puzicha
#t2009
#c239
#index413
#id424
#%
#!

#*mixture models mixture models interesting model family different uses mixture models include example generative component models clustering density estimation mixture models successfully used various kinds tasks modelling rate data clustering behaviour general modelling large heterogeneous populations step using mixture models define suitable model data parameter model estimated data phase called parameter estimation parameter learning important task finding number data supports problem called model learning paper discuss issues cover basic methods introduce modern methods numerous examples 
#@Petteri Nurmi
#t2009
#c50
#index414
#id426
#%
#!

#*view em algorithm incremental variants em algorithm performs maximum likelihood estimation data variables unobserved present function negative free energy step maximizes function respect model parameters step maximizes respect distribution unobserved variables perspective easy incremental variant em algorithm distribution unobserved variables 
#@Radford Neal, Geoffrey E. Hinton
#t2009
#c198
#index415
#id427
#%
#!

#*distributional clustering english words experimentally evaluate method automatically clustering words distribution particular syntactic words represented relative contexts appear relative entropy used measure clusters represented typical context distributions words according probabilities cluster membership cases 
#@Fernando C. N. Pereira, Naftali Tishby, Lillian Lee
#t2009
#c104
#index416
#id428
#%
#!

#*aggregate mixed order markov models statistical language processing consider use language models size accuracy intermediate different order gram models types models studied particular aggregate markov models models mapping words classes probabilistic mixed order markov models combine models predictions different words types models trained em algorithms maximum likelihood estimation examine smoothing procedures models different order grams significantly reduce perplexity unseen word 
#@Lawrence K. Saul, Fernando Pereira
#t2009
#c240
#index417
#id429
#%
#!

#*predicting performance combined ir systems introduce new technique modelsfi technique allows ir systems achieve using linear accurately predict performance combined based quantitative individual component systems taken trec applied linear model weighted sum technique supports hypotheses 
#@Christopher C. Vogt, Garrison W. Cottrell
#t2009
#c35
#index418
#id430
#%
#!

#*discovering objects location images given set images containing multiple object categories seek discover categories image locations achieve using generative models statistical text literature probabilistic latent semantic analysis plsa latent dirichlet allocation lda text analysis used discover topics corpus using bag words document representation discover topics object categories image containing instances categories mixture topics models applied images using visual word formed vector like region descriptors investigate set increasingly scenarios starting image sets containing object categories sets containing multiple categories including background object categories sample intra class scale variation categories approximate spatial layout demonstrate classification unseen images images containing multiple objects performance proposed unsupervised method compared approach 
#@Josef Sivic, Bryan C. Russell, Alexei A. Efros, Andrew Zisserman, William T. Freeman
#t2009
#c241
#index419
#id431
#%364,365,176,367,368,360,199,349,369,420,370,371,372,374,421,375,376,377,378
#!

#*affine invariant point paper presents novel approach detecting affine invariant method deal significant affine transformations including large scale transformations introduce significant changes point location scale shape approach allows solve problems based key ideas second matrix computed point used 
#@Krystian Mikolajczyk, Cordelia Schmid
#t2009
#c223
#index420
#id432
#%
#!

#*hierarchical dirichlet processes consider problems involving groups data observation group draw mixture model desirable share mixture components groups assume number mixture components unknown inferred data setting natural consider sets dirichlet processes group known clustering property dirichlet process provides nonparametric prior number mixture components group given mixture models various groups consider hierarchical model base measure dirichlet processes itself distributed according dirichlet process base measure discrete dirichlet processes share desired mixture models different groups share mixture components discuss representations hierarchical dirichlet processes terms process generalization chinese process refer chinese present markov chain monte carlo algorithms posterior inference hierarchical dirichlet process mixtures applications problems information retrieval text modelling 
#@Yee Whye Teh, Michael I. Jordan, Matthew J. Beal, David M. Blei
#t2009
#c133
#index421
#id433
#%422,423
#!

#*hierarchical bayesian models applications information retrieval article factors close factors achieve significant expected distribution words factors identify topics mixed random form document 
#@J. M. Bernardo, M. J. Bayarri, J. O. Berger, A. P. Dawid, D. Heckerman, A. F. M. Smith, M. West, David M. Blei, Michael I. Jordan, Andrew Y. Ng
#t2009
#c242
#index422
#id434
#%
#!

#*hidden markov model induction bayesian model merging paper describes technique learning number states topology hidden markov models examples induction process model consistent training data merging states choice states criterion bayesian posterior probability compare algorithm baum method estimating size models minimal hmms data cases estimation does converge requires parameters converge 
#@A Stolcke, S Omohundro
#t2009
#c239
#index423
#id435
#%560,561,562,563,564,565,566,567,568
#!

#*dynamic social network analysis using latent space models paper explores aspects social network modeling successful static model relationships dynamic model accounts time second make tractable learn models data number entities large generalized model entity point dimensional latent space points time large latent space observed links entities likely entities close latent space make model tractable sub quadratic number entities use appropriate functions similarity latent space use low dimensional trees new efficient dynamic adaptation multidimensional scaling approximate projection entities latent space efficient conjugate gradient update rule non linear local optimization time entity update log use synthetic real world data entities indicate near linear scaling computation time improved performance alternative approaches illustrate operating years data 
#@Purnamrita Sarkar, Andrew W. Moore
#t2009
#c190
#index424
#id436
#%425,426,427,428,429
#!

#*using domain knowledge facilitate automated association analysis link association analysis used intelligence domains extract search associations people large datasets link analysis challenging problems information high search complexity domain knowledge address challenges enable conduct automated effective efficient link analysis proposed techniques include concept space approach shortest path algorithm heuristic approach captures domain knowledge determining importance associations implemented called based proposed techniques results user study involving showed help subjects conduct link analysis efficiently subjects association paths based heuristic approach accurate based concept space approach 
#@Jennifer Schroeder, Jennifer Jie Xu, Hsinchun Chen
#t2009
#c243
#index425
#id437
#%
#!

#*clustering bipartite keyword graph paper present methods large bipartite approach employs flow based graph approach single link average link evaluate quality methods using additional textual compare results 
#@J. Carrasco, D. Fain, K. Lang, L. Zhukov
#t2009
#c244
#index426
#id438
#%
#!

#*collaboration analysis recommender systems using social networks researchers recommender use collaboration systems improves known important order information collaborative relationships recommender systems social paper measures collaboration 
#@Jordi Palau, Miquel Montaner, Beatriz Lopez, Josep Lluis de la Rosa
#t2009
#c245
#index427
#id439
#%
#!

#*modern multidimensional scaling book provides comprehensive multidimensional scaling family statistical techniques analyzing structure similarity data data widespread including example survey items direct ratings similarity choice objects trade indices set countries represents data points geometric space low dimensionality map help patterns data data matrices used model judgments similarity preference book used introduction students marketing particular background statistics book suited variety advanced topics mathematics required advanced topics developed systematically 
#@Ingwer Borg, Patrick Groenen
#t2009
#c30
#index428
#id440
#%
#!

#*optimization em expectation close relationship expectation maximization em algorithm direct optimization algorithms methods parameter learning identify conditions em like behavior conditions poor order convergence 
#@Ruslan Salakhutdinov, Sam Roweis, Zoubin Ghahramani
#t2009
#c246
#index429
#id441
#%
#!

#*dynamic topic models family probabilistic time series models developed analyze time evolution topics large document collections approach use state space models natural parameters multinomial distributions represent topics variational approximations based filters nonparametric regression developed approximate posterior inference latent topics addition quantitative predictive models sequential corpus dynamic topic models provide contents large document collection models demonstrated analyzing journal science 
#@David M. Blei, J Lafferty
#t2009
#c22
#index430
#id442
#%176,335,360,431,340,432,419,433,434,435
#!

#*new approach linear filtering prediction problems classical filtering prediction problem examined using representation random processes state transition method analysis dynamic systems new results formulation methods solution problem apply statistics growing memory filters nonlinear difference differential derived matrix optimal estimation error solution coefficients difference differential optimal linear filter obtained filtering problem shown dual noise free problem new method developed applied known problems extending earlier results discussion largely self contained principles basic concepts theory random processes 
#@R. E. Kalman
#t2009
#c247
#index431
#id443
#%
#!

#*author topic model authors documents introduce author topic model generative model documents latent dirichlet allocation lda include information author associated multinomial distribution topics topic associated multinomial distribution words document multiple authors modeled distribution topics mixture distributions associated authors apply model collection conference papers exact inference intractable datasets use sampling estimate topic author distributions compare performance generative models documents special cases author topic model lda topic model simple author model author associated distribution words distribution topics topics model demonstrate applications computing similarity authors entropy author output 
#@Michal Rosen-Zvi, Thomas L. Griffiths, Mark Steyvers, Padhraic Smyth
#t2009
#c129
#index432
#id444
#%176,444,361,360,199,445,446,440,205
#!

#*sparse gaussian processes using pseudo inputs present new gaussian process regression model locations pseudo input points learn gradient based number real data points obtain method training cost test joint method bayesian 
#@Edward Snelson, Zoubin Ghahramani
#t2009
#c190
#index433
#id445
#%
#!

#*nonparametric statistics text provides reader single book accounts number date issues nonparametric inference book level students statistics science engineering suitable researchers speed quickly modern nonparametric methods covers wide range topics including nonparametric method nonparametric regression density estimation orthogonal function methods estimation nonparametric confidence sets books dual approach includes mixture methodology theory 
#@L Wasserman
#t2009
#c30
#index434
#id446
#%
#!

#*bayesian dynamic models second edition book includes updated additional material structure theory application classes dynamic models bayesian time series analysis addition wide ranging updates central material second edition includes covers new topics research application bayesian 
#@M. West, Jeff Harrison
#t2009
#c30
#index435
#id447
#%
#!

#*author recipient topic model topic role discovery social networks experiments email previous work social network analysis modeled links entity language content topics links present author recipient topic art model social network analysis learns topic distributions based messages entities model latent dirichlet allocation author topic model adding key attribute distribution topics discovery topics according relationships people results email corpus researchers email providing evidence clearly relevant topics discovered art model better 
#@A McCallum, X Wang, A Corrada-Emmanuel
#t2009
#c248
#index436
#id448
#%437,176,438,439,440,205,441,421,442,443
#!

#*search social network address question small world experiment able short paths social network using local information immediate experiments network actual email organization social email network small world search strategies using position physical space organizational hierarchy relative target used line network data hierarchical structures defined local search strategies compare recent theoretical hypotheses underlying social structure en able simple search strategies discuss implications social software 
#@LA Adamic, Eytan Adar
#t2009
#c30
#index437
#id449
#%
#!

#*navigation small world small world phenomenon principle linked short chains investigated question feature range networks arising nature technology experimental study phenomenon fundamental components short chains second individuals operating local information finding chains issue investigate second modelling individuals short chains large social network page 
#@Jon M. Kleinberg
#t2009
#c249
#index438
#id450
#%
#!

#*structural equivalence individuals social networks problems structural attention researchers creation individual identity means social networks individuals embedded second structural equivalence problem non identification individuals means social networks aim article problems common framework identity equivalence individuals social networks simultaneously words study attempts structural identity equivalence just order achieve necessary approach represents social structure network social network individuals linked given relations concept place place analysis used 
#@F. lorrain, H. C. white
#t2009
#c250
#index439
#id451
#%
#!

#*multi text classification mixture model trained em important document classification tasks documents associated multiple class paper describes bayesian classification approach multiple classes document represented mixture labeled training data indicates classes generating document does indicate class generating use em missing value learning 
#@A McCallum
#t2009
#c251
#index440
#id452
#%
#!

#*probabilistic author topic models information discovery propose new unsupervised learning technique extracting information large text model documents generated stage stochastic author represented probability distribution topics topic represented probability distribution words words multi author paper assumed result mixture authors topic topic word author topic distributions learned data unsupervised manner using markov chain monte carlo apply methodology large corpus authors known digital library learn model discuss interpretation results discovered including specific topic author models ranking authors topic topics author significant trends science literature parsing topics authors detection papers specific online query interface model discussed allows interactive exploration author topic models corpora 
#@Mark Steyvers, Padhraic Smyth, Michal Rosen-Zvi, Thomas L. Griffiths
#t2009
#c3
#index441
#id453
#%
#!

#*identify search social networks social networks property people messages network specific target person steps present model social network terms personal defined number social dimensions model defines class networks method searching applicable network search problems including location data files peer peer networks pages world wide web information distributed databases 
#@D. J. Watts, P. S. Dodds, M. E. J. Newman
#t2009
#c252
#index442
#id454
#%
#!

#*information flow social groups present study information takes account observation item relevant person likely individuals social outside fact similarity node attributes social networks function graph distance model scale free network property finite threshold information limited tested predictions measuring messages organization numerical experiments consideration organizational distance individuals 
#@F Wu, BA Huberman, LA Adamic, JR Tyler
#t2009
#c253
#index443
#id455
#%
#!

#*missing link probabilistic model document content hypertext connectivity joint probabilistic model modeling contents inter connectivity document collections sets web pages research paper model based probabilistic factor decomposition allows identifying principal topics collection authoritative documents topics furthermore relationships topics order build predictive model link content applications approach information retrieval search topic identification query disambiguation focused web web analysis 
#@David Cohn, Thomas Hofmann
#t2009
#c254
#index444
#id456
#%
#!

#*new directions papers long used testing techniques despite fact similar paper value novel techniques applying problem techniques examined multivariate approach vocabulary analysis frequencies occurrence sets common high frequency words use machine learning based genetic algorithm seek relational expressions approaches produce results difficult problem 
#@D. I. HOLMES, R. S. FORSYTH
#t2009
#c255
#index445
#id457
#%
#!

#*digital libraries autonomous citation indexing world wide web way researchers access scientific articles increasingly available authors journal web sites online scientific information web largely article introduces creation digital libraries incorporating autonomous citation indexing fi citation indices similar science citation index fi 
#@S Lawrence, C. Lee Giles, Kurt D. Bollacker
#t2009
#c188
#index446
#id458
#%
#!

#*efficient distributed web information systems skyline queries place retrieval central databases application web information systems distributed aspect retrieval web sources variety nature information accessible internet extended query capabilities crucial efficiently perform distributed skyline queries essentially extend querying web information systems retrieval algorithm present useful heuristics speed retrieval practical cases challenges line information services discuss performance evaluations point open problems concept application modern information systems dimensionality problem skyline queries propose novel sampling scheme allows early skyline subsequent query refinement 
#@Wolf-Tilo Balke, Ulrich Guntzer, Jason Xin Zheng
#t2009
#c0
#index447
#id459
#%448,449,226,450,232,451,237,236,452,453,454,455,456,457
#!

#*situation aware mobile traffic information prototype mobile information services play important role future work private life areas needs tools novel techniques individual traffic planning navigation systems route planning difficult current information personal user preferences incorporated present prototype traffic information advanced personalized route planning including added services like traffic means fast changing nature data requires gather line internet sources given current bandwidth limitations update strategy central service database real time requirements providing information results personalized route planning query efficiently computed algorithm practical case studies technology automatically results various mobile devices summary evidence technologies preference modeling preference query evaluation advanced personalized mobile traffic information systems feasible today 
#@WT Balke, W Kiessling, C Unbehend
#t2009
#c256
#index448
#id460
#%
#!

#*fitness function multi objective planning fitness function used multi objective genetic algorithms obtain diverse set non fitness function derived definition properties explored modified fitness function proposed fitness functions compared state art fitness function literature results real world multi objective problem presented problem addresses use planning high growth regions 
#@BALLING Richard
#t2009
#c257
#index449
#id461
#%
#!

#*querying preferences user preferences increasingly important issue present information systems preferences used information filtering extraction reduce volume data presented user used track user policies improve decision making propose framework preferences relational query languages framework simple respect properties preferences makes possible different kinds preferences use preferences querying databases demonstrate framework numerous examples 
#@Jan Chomicki
#t2009
#c0
#index450
#id462
#%
#!

#*optimizing multi feature queries image databases digital libraries image retrieval queries similarity objects using attributes like shape texture color 
#@Ulrich Guntzer, Wolf-Tilo Balke, Werner Kiesling
#t2009
#c8
#index451
#id463
#%
#!

#*online algorithm skyline queries skyline queries ask set interesting points potentially large set data instance interesting better skyline queries retrieve interesting user choose promising paper present new online algorithm unlike existing algorithms compute skyline batch 
#@Donald Kossmann, Frank Ramsak, Steffen Rost
#t2009
#c8
#index452
#id464
#%
#!

#*preferences knowledge queries classical query languages provide way express data retrieved feature preferences desirable need preferences illustrated software engineering framework preference mechanism presented extension language domain relational family power resulting language discussed proposed mechanisms shown effectively allow use queries supporting software management functions 
#@M Lacroix, P Lavency
#t2009
#c258
#index453
#id465
#%
#!

#*efficient utility functions preferences equal preference provides convenient means constraints numeric utility functions direct numerical utility representations exponential cost paper describes efficient combine analysis utility independence search 
#@Michael McGeachie, Jon Doyle
#t2009
#c175
#index454
#id466
#%
#!

#*supporting ranked boolean similarity queries address needs applications require access retrieval multimedia objects developing multimedia analysis retrieval fi paper retrieval support content based queries image database content techniques extensively studied textual documents area automatic fi paper describes techniques 
#@Michael Ortega, Yong Rui, Kaushik Chakrabarti, Kriengkrai Porkaew, Sharad Mehrotra, T.Huang
#t2009
#c15
#index455
#id467
#%
#!

#*optimal algorithm skyline queries skyline set dimensional points contains points point dimensions skyline computation recently considerable attention database community especially online algorithms quickly return skyline points having read entire data file currently efficient algorithm nearest applies framework datasets indexed trees desirable features high speed initial skyline points arbitrary data distributions dimensions presents need elimination multiple accesses node large space overhead paper develop bound skyline algorithm based nearest neighbor search optimal performs single access tree nodes contain skyline points furthermore does retrieve space overhead significantly smaller finally simple implement efficiently applied variety alternative skyline queries experimental comparison shows outperforms usually orders magnitude problem instances 
#@Dimitris Papadias, Yufei Tao, Greg Fu, Bernhard Seeger
#t2009
#c10
#index456
#id468
#%
#!

#*advanced mobile services performing complex tasks web 
#@M. Wagner, WT Balke, M. Wagner, Wolf-Tilo Balke, R. Hirschfeld, W. Kellerer
#t2009
#c259
#index457
#id469
#%
#!

#*efficient document retrieval main memory disk access performance major traditional information retrieval systems compared memory disk bandwidth poor seek times problem considering query evaluation strategies main memory new techniques combined inverted list produce extremely high performance retrieval systems methods harm effectiveness evaluate techniques using new retrieval designed efficient query processing achieves improvement query previous methods 
#@Trevor Strohman, W. Bruce Croft
#t2009
#c35
#index458
#id470
#%459,460,461,37,462,42,463,232,464,45,46,465,466
#!

#*vector space ranking effective early termination considerable research effort improving effectiveness information retrieval systems techniques relevance feedback expansion provide better quality queries tested standard evaluation enhancements cost evaluating queries paper consider issue improve cost effectiveness searching new inverted file structure using weights provides superior retrieval effectiveness compared conventional inverted file structures early termination heuristics employed able similar effectiveness levels computational cost provide better cost performance previous inverted file 
#@Vo Ngoc Anh, Owen de Kretser, Alistair Moffat
#t2009
#c35
#index459
#id471
#%
#!

#*similarity scoring using term ranks propose method document ranking combines simple document view text fast evaluation strategies developed connection vector space model new method defines importance term document qualitatively reduces need parameters addition method supports fast query processing computation carried small integers dynamic pruning effective experiments wide range trec data new method provides retrieval effectiveness good better formulation variants language models 
#@Vo Ngoc Anh, Alistair Moffat
#t2009
#c35
#index460
#id472
#%
#!

#*query evaluation using pre computed exhaustive evaluation ranked queries expensive particularly small subset overall ranking required queries contain common terms gives rise techniques dynamic query pruning methods parts usual exhaustive evaluation generating good set answers query work propose new pruning methods make use impact sorted indexes compared exhaustive evaluation new methods reduce computation performed reduce memory required reduce data disk time allow performance guarantees terms precision mean average precision strong experiments using trec terabyte collection queries 
#@Vo Ngoc Anh, Alistair Moffat
#t2009
#c35
#index461
#id473
#%
#!

#*trec terabyte track trec year track track introduced trec single retrieval task trec track tasks named page finding task efficiency task tasks groups runs retrieval task groups runs named page finding task groups runs efficiency task report provides overview task results directions future background information development track track reports 
#@Stefan Buttcher, CLA Clarke, Ian Soboroff
#t2009
#c52
#index462
#id474
#%
#!

#*mapreduce data processing large clusters mapreduce programming model associated implementation processing generating large data sets users specify map function processes key value pair generate set intermediate key value pairs reduce function intermediate values associated intermediate key real world tasks model shown paper programs written functional style automatically large cluster machines run time takes details partitioning input data scheduling program execution set machines machine failures managing required inter machine communication allows experience parallel distributed systems easily utilize resources large distributed implementation mapreduce runs large cluster machines highly scalable typical mapreduce computation processes data thousands machines easy use hundreds mapreduce programs implemented mapreduce google clusters 
#@Jeffrey Dean, Sanjay Ghemawat
#t2009
#c7
#index463
#id475
#%467,468,469,470,471,472,473,474,475,476,477,478,479,480,481
#!

#*space limited ranked query evaluation using adaptive pruning evaluation ranked queries large text collections costly terms processing time memory space dynamic pruning techniques allow costs reduced potential risk retrieval effectiveness paper improved query pruning mechanism offers query evaluation costs retrieval effectiveness previous pruning approaches 
#@Nicholas Lester, Alistair Moffat, William Webber, Justin Zobel
#t2009
#c260
#index464
#id476
#%
#!

#*optimization strategies complex queries previous research efficiency text retrieval systems primarily methods consider inverted lists sequence methods known term time methods literature optimizing time systems sparse present improvement max score optimization efficient known document time scoring method like max score technique called term bounded max score guaranteed return scores documents evaluation particularly useful query model research technique explore problem space implemented large scale language modeling search engine tests corpus title queries method faster max score faster document time baseline optimized query times competitive conventional time systems years trec terabyte task 
#@Trevor Strohman, Howard R. Turtle, W. Bruce Croft
#t2009
#c35
#index465
#id477
#%482,63,483,484,45,61
#!

#*dbms cpu cache new execution engine improves execution speed main memory limitation introduces concept cache processing balance existing time execution primitives time model intermediate result large interpretation overhead respectively explain new query engine makes better use cache parallel computation resources modern orders magnitude faster commercial close hand programs computationally queries memory datasets address larger disk based datasets efficiency new storage layer developed bandwidth using compression cooperative 
#@Marcin Zukowski, Peter A. Boncz, Niels Nes, Sandor Heman
#t2009
#c261
#index466
#id478
#%
#!

#*high performance sorting networks workstations report performance sort collection sorting implementations network workstations parallel sorting competitive sorting large scale traditionally held performance records node cluster sort just node cluster benchmark implementations applied variety disk memory processor issues component evaluate use operating systems hardware parallel sorting existing primitives memory management file access aggregate communication disk bandwidth requirements 
#@Andrea C. Arpaci-Dusseau, Remzi H. Arpaci-Dusseau, David E. Culler, Joseph M. Hellerstein, David A. Patterson
#t2009
#c10
#index467
#id479
#%
#!

#*cluster making fast case common introduce data flow programming environment clusters computers designed provide maximum performance common case face non hardware software based simple design features high performance distributed storage redundancy mechanism called implemented number data applications design 
#@Remzi H. Arpaci-Dusseau, Eric Anderson, Noah Treuhaft, David E. Culler, Joseph M. Hellerstein, David A. Patterson, Katherine A. Yelick
#t2009
#c262
#index468
#id480
#%
#!

#*web world wide web potential used inexpensive convenient resource new challenges assumptions network workstations designed implemented set features commonly used network workstations user parallel program machine does account shared 
#@Arash Baratloo, Mehmet Karaul, Zvi M. Kedem, P. Wijckoff
#t2009
#c263
#index469
#id481
#%
#!

#*web search google cluster architecture application different queries run different processors partitioning overall index single query use multiple processors handle architecture features clusters fault software architecture achieves superior performance fraction cost built fewer expensive high end servers 
#@Luiz Andre Barroso, Jeffrey Dean, Urs Holzle
#t2009
#c264
#index470
#id482
#%
#!

#*explicit control batch aware distributed file present design implementation evaluation batch aware distributed file bad designed large batch workloads computing clusters distributed wide area bad consists novel components storage layer control traditionally fixed policies consistency replication control necessary different workloads extracting control storage layer external bad storage computation way dealing cache consistency fault space management issues specific manner using real workloads demonstrate performance explicit control excellent end end performance wide area 
#@John Bent, Douglas Thain, Andrea C. Arpaci-Dusseau, Remzi H. Arpaci-Dusseau, Miron Livny
#t2009
#c265
#index471
#id483
#%
#!

#*parallel operations study effects adding scan primitives unit time primitives parallel random access machine models presented shown primitives improve running time algorithms log factor greatly description algorithms significantly implement memory references algorithm free use operations memory reference author describes algorithms clearly illustrate scan primitives used algorithm design sort algorithm algorithm minimum tree algorithm line drawing algorithm merging algorithm run read addition scan primitives simpler efficient pure scan primitives implemented connection machine available parallel set machine 
#@Guy E. Blelloch
#t2009
#c183
#index472
#id484
#%
#!

#*cluster based scalable network services identify fundamental requirements scalable net work services incremental scalability growth availability fault cost effectiveness argue clusters high speed challenges internet server workloads pro software managing partial failures large cluster does new service end propose general building cluster based scalable network services requirements reuse service pro model based perform formation aggregation content performance implementation architecture programming model data semantics results ing consistency availability soft state robust management architecture used creating new allowing authors focus content service building blocks tion discuss real implementations services based web population search engine present performance based substantial evidence experience support architecture 
#@Armando Fox, Steven D. Gribble, Yatin Chawathe, Eric A. Brewer, Paul Gauthier
#t2009
#c266
#index473
#id485
#%
#!

#*systematic efficient scan list functions class introduced efficient parallel implementation functions class derived transformations directly arbitrary fixed number processors providing popular scan function parallel 
#@Sergei Gorlatch
#t2009
#c267
#index474
#id486
#%
#!

#*storage architecture early interactive search paper explores concept early search data processing data storage devices using code enables perform efficient application specific filtering large data collections early helps users items early possible consists set application generated filters uses object user evaluation order filters based run time filter computational cost partition computation storage host adjust changes hardware network conditions performance dynamically query run time state informal user study image retrieval application supports belief early significantly improves quality inter active searches 
#@Larry Huston, Rahul Sukthankar, Rajiv Wickremesinghe, Mahadev Satyanarayanan, Gregory R. Ganger, Erik Riedel, Anastassia Ailamaki
#t2009
#c268
#index475
#id487
#%
#!

#*parallel prefix computation prefix problem compute products fi operation construction used obtain product circuit solving prefix problem depth size bounded application yields fast small boolean state boolean depth size bounded obtained binary size significantly depth increase constant 
#@Richard E. Ladner, Michael J. Fischer
#t2009
#c112
#index476
#id488
#%
#!

#*efficient information security load balancing fault information algorithm developed file length pieces fi length fi pieces computationally efficient sum lengths fi chosen close space efficient numerous applications secure reliable storage information networks single fault efficient information networks communications processors parallel computers problem provably time efficient highly fault routing cube achieved using just constant size 
#@Michael Rabin
#t2009
#c269
#index477
#id489
#%
#!

#*active large scale data processing cost intelligence away cpu storage use trend computing power form complex processing storage date place relatively low levels storage trends storage density hardware hosts data propose using active disk storage combines processing memory soft allow level functions directly portions application processing storage significantly reduces data traffic present large systems reducing execution time basic data mining 
#@Erik Riedel, Christos Faloutsos, Garth A. Gibson, David Nagle
#t2009
#c188
#index478
#id490
#%
#!

#*distributed computing practice experience project users computing today project explore social technical problems cooperative computing scales ranging world wide computational grid chapter provide history project field distributed computing outline core components technology computing correspond social structures experience course research ideas grow systems 
#@Douglas Thain, Todd Tannenbaum, Miron Livny
#t2009
#c270
#index479
#id491
#%
#!

#*model parallel computation success model sequential computation fact efficient software hardware high level languages efficiently compiled model implemented hardware author software hardware required parallel computation widely used article introduces parallel model candidate role gives results efficiency implementing high level language features algorithms implemented hardware 
#@Leslie G. Valiant
#t2009
#c7
#index480
#id492
#%
#!

#*sort terabyte quickly node ibm sorted terabyte data byte records fi times faster previous record problem program itself designed benchmark cluster hardware disk operating file communication library software ibm aggregate data 
#@Jim Wyllie
#t2009
#c271
#index481
#id493
#%
#!

#*efficient query evaluation using level retrieval process present efficient query evaluation method based level approach level method parallel query term postings identifies candidate documents using approximate evaluation taking account partial information term occurrences query independent factors second level promising candidates fully evaluated exact scores computed efficiency evaluation process improved significantly using dynamic pruning techniques little cost effectiveness pruning user function time query evaluation experimentally using trec web track data determined algorithm significantly reduces total number evaluations loss precision recall heart approach efficient implementation new boolean construct called weak independent 
#@AZ Broder, David Carmel, Michael Herscovici, Aya Soffer, Jason Y. Zien
#t2009
#c32
#index482
#id494
#%485,486,487,488,63,489,490,491,492,66,493,67,68,494,495
#!

#*inquery retrieval larger heterogeneous text databases available information retrieval research depend development powerful efficient flexible retrieval engines paper retrieval inquery based probabilistic retrieval model provides support indexing complex query formulation inquery used successfully databases containing nearly documents introduction increasing 
#@James P. Callan, W. Bruce Croft, Stephen M. Harding
#t2009
#c272
#index483
#id495
#%
#!

#*combining language model inference network approaches retrieval inference network retrieval model implemented inquery search engine allows structured queries incorporates form ad hoc estimates word probabilities language modeling offers formal estimation techniques paper combine language modeling inference network approaches single framework resulting model allows structured queries evaluated using language modeling estimates explore issues involved combining smoothing proximity nodes experimental results presented comparing query likelihood model inquery new model results high quality structured queries outperform unstructured queries consistently achieves higher average precision inquery 
#@Donald Metzler, W. Bruce Croft
#t2009
#c63
#index484
#id496
#%
#!

#*fast match continuous speech recognition using models large vocabulary real time speech recognition need fast method selecting list candidate words vocabulary match given input paper highly accurate fast match continuous speech recognition algorithm uses models efficient search techniques select set candidate words models derived constructing decision trees query context given context models words vocabulary tree efficient tree search algorithms used select list candidate words using models using method able obtain accuracy fast continuous speech recognition task vocabulary words 
#@L.R. Bahl, Peter V. de Souza, P.S. Gopalakrishnan, D. Nahamoo, M.A. Picheny
#t2009
#c273
#index485
#id497
#%
#!

#*join operator text search consider problem efficiently sampling web search engine query results turn using small random sample instead set results leads efficient approximate algorithms applications determining set categories given search results finding range values associated result set order enable multi search estimating size result set data mining associations query terms present analyze efficient algorithm obtaining uniform random samples applicable search engine based posting lists document time evaluation knowledge popular web search engines google class furthermore algorithm modified modern object oriented approach posting lists viewed streams method method boolean complex queries built method terms case construct basic method samples term posting lists probability construct methods boolean operators methods finally test efficiency quality approach synthetic real world data 
#@K. Beyer, A. Jhingran, B. Lyle, S. Rajagopalan, E. Shekita
#t2009
#c50
#index486
#id498
#%
#!

#*space time trade hash coding errors paper trade certain computational factors hash coding analyzed paradigm problem considered testing series messages membership given set messages new hash coding methods examined compared particular conventional hash coding method computational factors considered size hash area space time required identify message given set time error frequency new methods intended reduce space required contain hash information associated conventional methods reduction space exploiting small fraction errors applications particular applications large data involved core hash area feasible using conventional methods applications overall performance improved using smaller core hash area new methods necessary using time consuming test small fraction errors associated new methods example discussed possible areas application new methods analysis paradigm problem demonstrates allowing small number test messages identified given set smaller hash area used increasing time 
#@Burton H. Bloom
#t2009
#c7
#index487
#id499
#%
#!

#*large scale web search engine paper present google prototype large scale search engine makes use structure present hypertext google designed index web efficiently produce satisfying search results existing systems prototype text database million pages available search engine challenging task search engines index hundreds millions web pages involving comparable number terms answer queries despite importance large scale search engines web little research furthermore rapid advance technology web creating web search engine today different years paper provides depth description large scale web search engine detailed public description know date problems scaling traditional data magnitude new involved using additional information present hypertext produce better search results paper addresses question build practical large scale exploit additional present hypertext look problem effectively deal hypertext collections 
#@Sergey Brin, Lawrence Page
#t2009
#c46
#index488
#id500
#%
#!

#*object oriented interface index implemented method searches index records database using object oriented interface record unique address database index index entries index entry including word ordered list locations word database words represent unique information database index entries ordered according order words second according order locations associated word query terms operators term associated corresponding index entry operators terms basic stream reader object generated term query basic stream reader object sequentially locations corresponding index entry determine target location compound stream reader object generated operator compound stream reader object references basic stream reader objects associated terms related operator compound stream reader object produces locations words single record according operator basic compound stream reader object data references method references operate data references 
#@Burrows, Michael
#t2009
#c274
#index489
#id501
#%
#!

#*method parsing indexing searching world wide web pages indexes web pages internet pages stored computers connected communications network page unique record pages include links pages communication interface connected internet used batch web pages computers links determined automated web browser connected communications interface parser sequentially partitions batch specified pages words word represents portion information specific page word represents attribute portions specific page parser sequentially assigns locations words locations indicates unique occurrences word web output parser stored memory index index includes index entry unique word index entry includes location entries unique word web query module query terms operators operators terms search engine uses object oriented stream readers sequentially read location specified index entries specified index entries correspond terms query module presents pages search engine users web 
#@Patent
#t2009
#c19
#index490
#id502
#%
#!

#*trec experiments index pruning year ir developed past years ibm research web track experiments focused ad hoc tasks main goal experiments novel pruning method presented significantly reduces size index little influence systems precision reducing index size feasible index large text collections web tracks data low end machines furthermore using method web search engines significantly decrease storing extremely large indices entries influence search results 
#@David Carmel, Einat Amitay, Miki Herscovici, Yoelle S. Maarek, Yael Petruschka, Aya Soffer
#t2009
#c52
#index491
#id503
#%
#!

#*shortest ranking address trec topics used precise query language yields combines arbitrary text pre defined like words documents solution inverse length shortest containing document sum scores solutions strategy documents documents satisfying queries added lower rank results 
#@CLA Clarke, Gordon V. Cormack, Forbes J. Burkowski
#t2009
#c275
#index492
#id504
#%
#!

#*term ordered query evaluation versus document ordered query evaluation large document databases main technique eficient processing ranked queries large text collections document ordered processing term ordered processing compare techniques experimentally similar costs short queries long queries document ordered processing costly overall conclude term ordered processing limited hierarchical index eficient mechanism 
#@Marcin Kaszkiel, Justin Zobel
#t2009
#c35
#index493
#id505
#%
#!

#*optimal distributed database systems filter based algorithm distributed database systems algorithm reduces communications costs process distributed natural join possible filter optimal filter developed filter information used effective process quickly effective use 
#@James K. Mullin
#t2009
#c276
#index494
#id506
#%
#!

#*overview text retrieval conference trec trec large web task based million page corpus tasks topic relevance task like traditional trec ad hoc used queries taken real web search logs description topic description inferred topic topics task queries entity home 
#@EM Voorhees, Donna Harman
#t2009
#c277
#index495
#id507
#%
#!

#*evaluating queries accessible databases global page ordering query web search engine usually consists list keywords search engine best pages query query model multimedia collections general relational data certain applications example consider relation information available including location price range overall rating user queries relation simply specify users location target price range return best terms combination proximity user closeness match target price range overall rating processing queries efficiently challenging number reasons critical web applications relation attributes available external web accessible form interfaces query potentially large set candidate objects paper study process queries efficiently setting attributes users specify target values external autonomous sources variety access interfaces present algorithms processing queries evaluate using synthetic real web accessible data 
#@Amelie Marian
#t2009
#c278
#index496
#id508
#%227,228,497,498,499,500,232,501,451,502,503,504,505,506,455,507
#!

#*optimizing queries multimedia repositories repositories multimedia objects having multiple types attributes image text increasingly common selection attributes typically produce just set objects traditional relational query model filtering grade match associated object object matches selection condition ranking multimedia repositories allow access attributes object indexes investigate optimize processing queries multimedia repositories key issue choice indexes used search define execution space search minimal set indexes minimal general problem optimal plan search minimal execution space np hard solve problem efficiently predicates query independent problem optimizing queries ask ranked objects viewed cases evaluating selection conditions problems viewed extended filtering problem 
#@Surajit Chaudhuri, Luis Gravano
#t2009
#c10
#index497
#id509
#%
#!

#*evaluating selection queries applications users specify target values certain attributes requiring exact matches values instead result queries typically rank best match given attribute paper study advantages limitations processing query single range query traditional relational process particular study determine range query 
#@Surajit Chaudhuri, Luis Gravano
#t2009
#c8
#index498
#id510
#%
#!

#*probabilistic optimization queries problem finding best answers query quickly finding answers increasing importance relational databases applied multimedia decision support domains approach efficiently answering queries augment query additional selection away portion answer set risk selection fewer desired number answers execution 
#@Donko Donjerkovic, Raghu Ramakrishnan
#t2009
#c8
#index499
#id511
#%
#!

#*combining fuzzy information multiple systems traditional database result query set values values satisfy query data servers queries based image content text retrieval systems result query sorted list example case queries based image content query ask objects particular red result query sorted list objects database sorted 
#@Ronald Fagin
#t2009
#c137
#index500
#id512
#%
#!

#*practical approach combined querying databases web present disk new approach combining query traditional databases existing search engines web supported database queries results web searches enhance sql queries relational database database supported web queries uses information stored database enhance paper focuses primarily describing simple low overhead way 
#@Roy Goldman, Jennifer Widom
#t2009
#c10
#index501
#id513
#%
#!

#*optimizing queries expensive predicates traditional focus relational query optimization schemes choice join methods join orders restrictions typically query rules apply restrictions random order joins possible rules work assumption restrictions essentially zero time operation today object oriented database systems allow users define time consuming functions used query join predicates furthermore sql long supported predicates time consuming restrictions considered zero time operations model query optimization enhanced paper develop theory expensive predicates query plan total cost plan including costs joins restrictions minimal present algorithm implement theory results implementation experience enhanced orders magnitude faster plans generated traditional query additional complexity considering expensive predicates optimization small 
#@Joseph M. Hellerstein, Michael Stonebraker
#t2009
#c10
#index502
#id514
#%
#!

#*prefer efficient execution multi parametric ranked queries users need optimize selection objects weighting importance multiple object optimization problems appear research applied mathematics select home weighted number attributes like distance area capture queries definition use weight function relations derive score database systems produce results preference query need evaluate weight function relation prefer answers preference using materialized views pre stored result preference query pipelined fashion using materialized view excellent performance reasonable number materialized views pro algorithm selects number views given space constraints implemented algorithms proposed prototype called prefer commercial database management results performance comparison ing algorithms prior approaches using results indicate proposed superior performance compared approaches materialized views execution time 
#@Vagelis Hristidis, Nick Koudas, Yannis Papakonstantinou
#t2009
#c10
#index503
#id515
#%
#!

#*optimizing queries expensive predicates work propose assess technique called optimizing evaluation expensive technique optimizing selection predicates evaluation costs vary evaluation nested defined function object oriented extended orders magnitude attribute access 
#@Alfons Kemper, Guido Moerkotte, Klaus Peithner, Michael Steinbrunn
#t2009
#c10
#index504
#id516
#%
#!

#*incremental join queries ranked inputs paper problem incremental joins multiple ranked data sets join condition list arbitrary user defined predicates input problem arises important applications dealing ordered inputs multiple ranked data sets requiring solutions use multimedia applications examples problem equally applicable traditional database applications involving optimal resource allocation scheduling decision making ranking propose enables querying ordered data sets arbitrary join predicates basic version algorithm does use random access variation exploit available indexes efficient random access based join predicates special case includes join considered fagin joins based identical case algorithms perform efficiently main contribution generalization join scenarios previously including cases random access algorithm possible unique addition support multiple join levels nested join modeling multimedia data approximation versions algorithms finally strong results proposed algorithms study performance empirically 
#@Apostol Natsev, Yuan-Chi Chang, J.R. Smith, Chung-Sheng Li, Jeffrey Scott Vitter
#t2009
#c8
#index505
#id517
#%
#!

#*query processing issues image multimedia databases multimedia databases systems content retrieval ibm essential effectively efficiently use existing large collections image data modern aim systems images based database 
#@Surya Nepal, M. V. Ramakrishna
#t2009
#c55
#index506
#id518
#%
#!

#*numerical art scientific computing book like edition methods numerical computing practical efficient possible book reader particular tasks view particularly large guide main lead practical book results decision literature computation discuss possible method practical relative offer practical judgments gain experience form opinion reliable able read programs language version numerical second edition book numerical second edition separately available prefer program language earlier numerical numerical examples basic available containing additional material second edition versions versions basic language choice 
#@William H. Press
#t2009
#c30
#index507
#id519
#%
#!

#*exploiting hidden sentiment links reviews past years research effort opinion mining product reviews recently research feature level opinion mining instead review level identifying semantic product feature opinion words context methods usually utilize explicit product feature words opinion words reviews approach based explicit loss hidden sentiment association product features opinion words paper propose novel unsupervised mutual approach deal feature level product opinion mining problem approach clusters product features opinion words simultaneously content information sentiment link information framework based product feature categories opinion word groups construct sentiment association set groups data objects identifying sentiment links knowledge multi source incorporated enhance clustering procedure based pre constructed association set approach largely predict opinions different product features case explicit product feature words reviews provides accurate opinion evaluation experimental results demonstrate method outperforms state art algorithms 
#@Qi Su, Xinying Xu, Honglei Guo, Zhili Guo, Xian Wu, Xiaoxun Zhang, Bin Swen, Zhong Su
#t2009
#c38
#index508
#id520
#%509,510,511,512,513,385,399,142,514,144,515,146,516,517,388,518,519,150,520
#!

#*evaluating measures lexical semantic lexical semantic applications nlp different measures proposed evaluate measures use wordnet central resource comparing performance detecting correcting real word spelling errors information content based measure proposed superior proposed lin addition explain distributional similarity lexical semantic 
#@Alexander Budanitsky, Graeme Hirst, Er Budanitsky Graeme Hirst
#t2009
#c279
#index509
#id521
#%
#!

#*noun phrase coreference clustering paper introduces new unsupervised algorithm noun phrase coreference resolution existing methods views resolution clustering task evaluation coreference resolution corpus algorithm achieves measure worst best systems evaluation clustering approach outperforms treat coreference resolution learning problem clustering algorithm appears provide flexible mechanism application context independent context dependent constraints preferences accurate partitioning noun phrases coreference equivalence classes 
#@Claire Cardie, Kiri Wagstaff
#t2009
#c280
#index510
#id522
#%
#!

#*mining opinion extraction semantic classication product reviews web contains wealth product reviews task opinion mining tool process set search results given item generating list product attributes quality features opinions poor mixed good begin identifying unique properties problem develop method automatically distinguishing positive negative reviews classifier information retrieval techniques feature extraction scoring results various metrics heuristics vary depending testing situation best methods work better traditional machine learning operating individual sentences collected web searches performance limited noise context complete web based tool simple method grouping sentences attributes results qualitatively quite useful 
#@K Dave, S Lawrence, DM Pennock
#t2009
#c281
#index511
#id523
#%
#!

#*summarizing arguments subjective documents supporting decision making world wide web volume subjective information opinions reviews increasing rapidly trends rules latent large set subjective potentially useful decision making purposes paper propose method summarizing subjective specifically opinions pro arguments target topic introduce users summarize arguments topic order choose reasonable decision making evaluate called experimentally 
#@Atsushi Fujii, Tetsuya Ishikawa
#t2009
#c110
#index512
#id524
#%
#!

#*chinese named entity recognition based multilevel linguistic features paper presents chinese named entity recognition employs robust risk classification method incorporates advantages based word based models experiments large scale corpus significant performance enhancements obtained integrating various linguistic information chinese word segmentation semantic types speech named entity basic chinese based model novel feature weighting mechanism employed obtain useful important linguistic features limitation computational resources building high quality named entity recognition large scale corpus samples selected active learning approach 
#@Honglei Guo, Jianmin Jiang, Gang Hu, Tong Zhang
#t2009
#c282
#index513
#id525
#%
#!

#*methods classication analysis multivariate observations paper describes number applications means procedure classifying random sample points sub procedure consists starting groups single random point adding points group mean point nearest point added group mean group account new point stage fact means group sample way points classified basis final means portions result tend sense having low class variance applications suggested problems non linear prediction efficient communication non parametric tests independence similarity grouping automatic file construction extension methods general metric spaces 
#@James B.MacQueen
#t2009
#c283
#index514
#id526
#%
#!

#*discovering word senses text manually compiled usually serve source word senses include rare senses missing corpus domain specific senses present clustering algorithm called clustering automatically word senses text set clusters called similarity space used feature vector cluster assigning words similar clusters assigning element cluster features element allows discover frequent senses word avoid discovering senses cluster word represents senses present evaluation methodology automatically measuring precision recall discovered senses 
#@Patrick Pantel, Dekang Lin
#t2009
#c3
#index515
#id527
#%
#!

#*comparative web search paper define study novel search problem comparative web search task seek relevant comparative information web help users conduct comparisons set topics called developed effectively facilitate web users comparison needs given set queries represent topics user wants compare characterized automatic retrieval ranking web pages incorporating relevance queries comparative contents contain automatic clustering comparative contents semantically meaningful themes extraction representative summarize differences comparative contents theme developed novel interface supports types view pair view result page level cluster view comparative pages themes extracted phrases facilitate users comparison experiment results effective efficient 
#@Jian-Tao Sun, X Wang, Dou Shen, Hua-Jun Zeng, Zheng Chen
#t2009
#c38
#index516
#id528
#%
#!

#*elements information theory essential topics information theory including entropy data compression channel rate distortion network information theory hypothesis testing authors provide readers understanding underlying theory applications problem sets summary end chapter readers historical chapter main points 
#@TM Cover, Joy A. Thomas
#t2009
#c30
#index517
#id529
#%
#!

#*constrained means clustering background knowledge clustering traditionally viewed unsupervised method data analysis cases information problem domain available addition data instances paper demonstrate popular means clustering algorithm modified make use information experiments artificial constraints data sets observe improvements clustering accuracy apply method real world problem automatically detecting data observe increases performance 
#@Kiri Wagstaff, Claire Cardie, Seth Rogers, Stefan Schrodl
#t2009
#c22
#index518
#id530
#%
#!

#*web object categorization existing categorization algorithms deal homogeneous web objects consider objects additional features taking types objects account focusing single aspect inter object relationship sufficient fully reveal true categories web objects paper propose novel categorization algorithm called iterative categorization algorithm exploit different types web objects web including web pages queries web objects individual classification results different types objects experiments clickthrough log dataset search engine terms measure achieves improvement pure content based classification method achieves improvement query based method improvement measure known virtual document based method experiments fast applicable real world applications 
#@Gui-Rong Xue, Yong Yu, Dou Shen, Qiang Yang, Hua-Jun Zeng, Zheng Chen
#t2009
#c113
#index519
#id531
#%
#!

#*framework clustering heterogeneous web objects paper introduce novel framework clustering web data heterogeneous nature existing methods integrate heterogeneous data unified feature space explore adjust effect different heterogeneous information contrast framework enables separate clustering homogeneous data entire process based features structure link information 
#@Hua-Jun Zeng, Zheng Chen, Wei-Ying Ma
#t2009
#c260
#index520
#id532
#%
#!

#*fast robust fixed point algorithms independent component analysis independent component analysis ica statistical method observed random vector components statistically independent possible paper use combination different approaches linear ica information theoretic approach projection pursuit approach using maximum entropy approximations differential en introduce family new contrast objective functions ica contrast functions enable estimation decomposition mutual information estimation independent components projection pursuit directions statistical properties estimators based contrast functions analyzed assumption linear mixture model shown choose contrast functions robust minimum variance finally simple point algorithms practical optimization contrast functions algorithms optimize contrast functions fast 
#@Aapo Hyvarinen
#t2009
#c284
#index521
#id533
#%522,523,524,525,517,526,527,528,529,530,531,532,533,534,535,536,537,538
#!

#*new learning algorithm blind source separation new line learning algorithm minimizes statistical outputs derived blind separation mixed signals measured average mutual information mi outputs source signals matrix unknown number sources gram expansion instead expansion used evaluating mi natural gradient approach used minimize mi novel activation function 
#@S. Amari, A. Cichocki, H. H. Yang
#t2009
#c239
#index522
#id534
#%
#!

#*information maximization approach blind separation blind derive new self organizing learning algorithm maximizes information network nonlinear algorithm does assume knowledge input distributions defined zero noise limit conditions information maximization properties linear case transfer function able higher order input distributions perform true redundancy reduction output representation enables network separate statistically independent components inputs higher order generalization principal components analysis apply network source separation party problem successfully separating unknown mixtures variant network architecture able perform blind unknown speech signal finally derive dependencies information transfer time suggest information maximization provides unifying framework problems blind signal processing 
#@Anthony J. Bell, Terrence J. Sejnowski
#t2009
#c285
#index523
#id535
#%
#!

#*adaptive source separation source separation consists set independent signals mixtures unknown coefficients observed paper introduces class adaptive algorithms source separation adaptive version estimation called adaptive separation independence algorithms based idea updating specific form matrix updates systematically yields algorithms simple structure real complex mixtures performance algorithm does depend matrix particular convergence rates conditions levels depend normalized distributions source signals closed form expressions quantities given performance analysis theme paper source separation problem underlying structure parameter space forms matrix group explore fact implementation performance optimization algorithms 
#@JF Cardoso, BH Laheld
#t2009
#c286
#index524
#id536
#%
#!

#*independent component new concept signal processing applying unsupervised learning techniques like ica temporal key question discovered reliable words error assess quality separation use methods questions experimentally proposed variance strongly correlated separation error demonstrate estimation used choose appropriate 
#@P. COMON
#t2009
#c287
#index525
#id537
#%
#!

#*projection pursuit algorithm exploratory data analysis algorithm analysis multivariate data presented discussed terms specific examples algorithm seeks dimensional linear multivariate data relatively highly 
#@JH Friedman, J. W. Tukey
#t2009
#c288
#index526
#id538
#%
#!

#*exploratory projection pursuit exploratory projection pursuit tool exploratory data analysis theorem allows data work search non normal density index functions used structure projection local maximum index function provides interesting dimensional views data variety index functions 
#@JH Friedman
#t2009
#c289
#index527
#id539
#%
#!

#*experimental comparison neural ica algorithms neural algorithms independent component analysis ica introduced computational properties systematically studied paper compare accuracy convergence speed computational load properties neural semi neural ica algorithms comparison interesting differences algorithms introduction independent component analysis ica unsupervised technique 
#@X Giannakopoulos, J Karhunen, E Oja
#t2009
#c290
#index528
#id540
#%
#!

#*robust statistics statistics classical methods rely assumptions practice particular assumed data distributed approximately central limit theorem produce distributed estimates unfortunately outliers data classical methods poor performance robust statistics seeks provide methods classical methods affected outliers small model assumptions 
#@PJ Huber, EM Ronchetti
#t2009
#c30
#index529
#id541
#%
#!

#*modern factor analysis edition authoritative text incorporates new advances science technology years author gives coverage theoretical applied aspects factor analysis foundations advanced techniques highly text researchers students working statistics related 
#@Harry H. Harman
#t2009
#c30
#index530
#id542
#%
#!

#*projection pursuit article concerns automatic used non linear structures clusters general phrase non linear structures refers structures easily using sample mean sample summary statistics like sample mean reveal linear structures scale correlation structures data 
#@PJ Huber
#t2009
#c291
#index531
#id543
#%
#!

#*unit contrast functions independent component analysis statistical analysis author introduced previously large family unit contrast functions used independent component analysis ica paper family analyzed case sample aspects estimators obtained using contrast functions considered variance robustness outliers expression contrast function minimizes variance obtained function probability 
#@Aapo Hyvarinen
#t2009
#c292
#index532
#id544
#%
#!

#*new approximations differential entropy independent component analysis projection pursuit derive order approximation density maximum entropy continuous random variable given number simple constraints results density expansion similar classical polynomial density gram using approximation density approximation differential entropy derived approximation entropy exact robust outliers classical approximation 
#@Aapo Hyvarinen
#t2009
#c190
#index533
#id545
#%
#!

#*fast algorithm estimating ica image introduce fast method estimating independent components image data based concept means high dimensional space large set vectors orthogonal estimate basis using ica algorithms partial different independent components method implemented using algorithm leads computationally highly efficient method 
#@Aapo Hyvarinen, Razvan Cristescu, E Oja
#t2009
#c293
#index534
#id546
#%
#!

#*independent component analysis general nonlinear hebbian like learning rules number neural learning rules recently proposed paper fact ica performed simple hebbian hebbian learning rules weak relations information theoretical quantities practically non linear function used learning rule provided hebbian hebbian term chosen addition hebbian like mechanism weight vector 
#@Aapo Hyvarinen, E Oja
#t2009
#c294
#index535
#id547
#%
#!

#*blind separation sources blind source separation problem extract underlying source signals set linear mixtures matrix unknown situation common radio signal image processing areas suggest stage separation process selection possibly signal dictionary instance learned dictionary sources assumed followed sources exploiting sparse consider general case sources mixtures derive efficient algorithm case dictionary equal numbers sources mixtures experiments artificial signals demonstrate significantly better separation known techniques 
#@P. COMON, C. JUTTEN, J. HERAULT
#t2009
#c294
#index536
#id548
#%
#!

#*class neural networks independent component analysis independent component analysis ica recently developed useful extension standard principal component analysis pca ica model blind separation unknown source signals linear mixtures application source signals correspond coefficients ica expansion paper propose neural structures related networks performing complete ica basic ica network consists separation basis vector estimation used blind source separation estimation basis vectors ica consider learning algorithms layer previous nonlinear pca type algorithms separation capabilities greatly improved proposed class networks yields good results test examples artificial real world data 
#@J Karhunen, E Oja, Liuyue Wang, Ricardo Vigario, Jyrki Joutsensalo
#t2009
#c295
#index537
#id549
#%
#!

#*extraction artifacts eeg using independent component analysis information science university technology fi activity main sources eeg new approach presented using statistical technique independent component analysis technique components distribution time distinguishing signals occurring signals occurring signals category usually formed approach possible pure activity eeg including reduce brain activity extracting portions signals 
#@RN Vigrio
#t2009
#c296
#index538
#id550
#%
#!

#*gap factor model discrete data present probabilistic model document corpus combines desirable features previous models model called gap poisson distributions random variable gap factor model gives approximate factorization document term matrix product matrices factors non negative terms gap generative probabilistic model assigns finite probabilities documents corpus computed efficient simple em suitable choice parameters gap factorization maximizes independence factors used independent component algorithm document data form gap model empirically gives accurate results probabilistic model measured perplexity retrieval model gap model documents terms low dimensional space themes models texts terms theme categories subject descriptors 
#@John F. Canny
#t2009
#c35
#index539
#id551
#%176,176,199,521,540,541,542,543,352,544,545
#!

#*independent component analysis algorithms applications fundamental problem neural network research finding suitable representation multivariate data random vectors reasons computational representation linear transformation original data words component representation linear combination original variables known linear transformation methods include principal component analysis factor analysis projection pursuit independent component analysis ica recently developed method goal linear representation non gaussian data components statistically independent independent possible representation capture essential structure data applications including feature extraction signal separation paper present basic theory applications ica recent work subject 
#@Aapo Hyvarinen, E Oja
#t2009
#c184
#index540
#id552
#%
#!

#*solution problem latent semantic analysis theory induction representation knowledge people know little information problem takes forms learning vocabulary text especially convenient case research new general theory acquired similarity knowledge representation latent semantic analysis lsa presented used successfully learning phenomena global knowledge local occurrence data large representative text lsa acquired knowledge vocabulary english comparable rate lsa uses prior linguistic similarity knowledge based general mathematical learning method achieves powerful inductive effects extracting right number dimensions represent objects contexts relations phenomena problems 
#@TK Landauer, S. Dumais
#t2009
#c205
#index541
#id553
#%
#!

#*introduction latent semantic analysis latent semantic analysis lsa theory method extracting representing contextual usage meaning words statistical computations applied large corpus text underlying idea aggregate word contexts given word does does appear provides set mutual constraints largely determines similarity meaning words sets words human knowledge established variety ways example scores overlap humans standard vocabulary subject matter tests human word sorting category judgments lexical data reported following articles issue accurately estimates individual students quality knowledge contained 
#@TK Landauer, PW Foltz, D Laham
#t2009
#c297
#index542
#id554
#%
#!

#*learning human like knowledge singular value decomposition report singular value decomposition viewed method unsupervised training network classes events linear connections single hidden layer used learn represent relations large numbers words large numbers natural text result dimensional semantic spaces trained added word represented vector similarities measured contained vectors good accuracy human judgments demonstrated performance multiple choice vocabulary domain knowledge tests expert evaluations ways examples given knowledge extracted method applied 
#@Thomas K. Landauer, D Laham, PW Foltz
#t2009
#c190
#index543
#id555
#%
#!

#*document clustering based non negative matrix factorization paper propose novel document clustering method based non negative factorization term document matrix given document corpus latent semantic space derived non negative matrix factorization captures base topic particular document cluster document represented combination base topics cluster membership document easily determined finding base topic document largest projection value experimental evaluations proposed document clustering method latent semantic indexing spectral clustering methods easy reliable document clustering results document clustering 
#@Wei Xu, Xin Liu, Yihong Gong
#t2009
#c35
#index544
#id556
#%
#!

#*study smoothing methods language models applied ad hoc information retrieval language modeling approaches information retrieval promising problem retrieval language model estimation studied extensively application areas speech recognition basic idea approaches estimate language model document rank documents likelihood query according estimated language model core problem language model estimation smoothing maximum likelihood estimator correct data paper study problem language model smoothing influence retrieval performance examine retrieval performance smoothing parameters compare popular smoothing methods different test collections 
#@ChengXiang Zhai, J Lafferty
#t2009
#c35
#index545
#id557
#%
#!

#*group topic discovery relations text present probabilistic generative model entity relationships textual attributes simultaneously groups entities topics corresponding text block models relationship data studied social network analysis time simultaneously cluster incorporating words associated certain relationships significantly joint inference allows discovery groups topics present experimental results large data sets years corresponding text voting records years similar data comparison traditional separate latent variable models words group topic models joint inference improves groups topics discovered categories subject descriptors 
#@Xuerui Wang, Natasha Mohanty, A McCallum
#t2009
#c298
#index546
#id558
#%547,548,549,176,550,551,552,553,554,555,556,557,558,559
#!

#*clustering search engine query log paper introduces technique mining collection transactions internet search engine similar queries similar exploit clickthrough data record user query search engine user selected candidates search engine dataset vertices corresponding 
#@Doug Beeferman, Adam L. Berger
#t2009
#c3
#index547
#id559
#%
#!

#*multi way distributional clustering interactions present novel unsupervised learning scheme simultaneously clusters variables types documents words authors based interactions types observed occurrence data scheme multiple clustering systems generated objective function measures multiple mutual information cluster variables implement idea pro algorithm clustering variables clustering variables local optimization focusing document clustering present extensive empirical study way way way applications scheme using real world datasets including news 
#@Ron Bekkerman, Ran El-Yaniv, A McCallum
#t2009
#c22
#index548
#id560
#%
#!

#*group detection using links categories subject descriptors 
#@Indrajit Bhattacharya, Lise Getoor
#t2009
#c299
#index549
#id561
#%
#!

#*information theoretic clustering dimensional occurrence tables arise frequently important applications text web log market data basic problem table analysis clustering clustering novel theoretical formulation views table empirical joint probability distribution discrete random variables clustering problem optimization problem information theory 
#@Inderjit S. Dhillon, Subramanyam Mallela, Dharmendra S. Modha
#t2009
#c3
#index550
#id562
#%
#!

#*does make connections countries investigate complex relationships countries past voting data terms network despite analysis shows european countries equally finding core countries significantly addition analysis enables widely held belief countries expected ones proximity complexity group self process central complexity representative real world situations groups agents establish inter relationships decide possible examples include groups individuals political groups 
#@Daniel Fenn, Omer Suleman, Janet Efstathiou, Neil F. Johnson
#t2009
#c300
#index551
#id563
#%
#!

#*power parties cohesion european parliament political parties european parliament form answers questions central understanding impact european parliament european policies questions central study behaviour general collected total population european parliament end data growing party cohesion despite growing internal european party groups distance parties left right dimension predictor patterns increased power european parliament increased power parties increased internal party cohesion inter party 
#@Hix Simon, Abdul Noury, Grard Roland
#t2009
#c301
#index552
#id564
#%
#!

#*analyzing similarities networks clusters analyze year employed methods used science community analysis text information theoretic measures assess association pairs based furthermore evaluate influence information channel matrix associations using hierarchical clustering multi dimensional scaling link analysis discrete latent variable model identify contrast continuous ideal point methods voting model weighted able estimate empirical voting power individual analysis 
#@A Jakulin, W Buntine, Timothy La Pira, Holly Brasher
#t2009
#c302
#index553
#id565
#%
#!

#*discovering latent classes relational data present framework learning relational knowledge people physical approach based generative relational model latent classes simultaneously determines kinds entities exist domain latent classes relations classes possible likely model previous models category learning attributes associated individual categories relationships apply domain general framework specific problems learn ing structure systems learning 
#@Kemp Charles, Thomas L. Griffiths, Joshua B. Tenenbaum
#t2009
#c50
#index554
#id566
#%
#!

#*model structure organization present network based approach characterize architectures terms domain elements individuals tasks resources possible relations elements results relational primitives resources assignment individuals tasks networks relations resources demonstrate utility model organizational theory generating series hypotheses performance 
#@KM Carley, D Krackhardt
#t2009
#c303
#index555
#id567
#%
#!

#*stochastic link group detection link detection analysis long important social sciences intelligence significant effort focused structural functional analysis known detection individual links important usually techniques result known recently internet sources led data provide probabilistic evidence occurrence news 
#@Jeremy Kubica, Andrew W. Moore, Jeff G. Schneider, Yiming Yang
#t2009
#c175
#index556
#id568
#%
#!

#*topic role discovery social networks previous work social network analysis modeled links en language content topics links present author recipient topic art model social network analysis learns topic distributions based direction sensitive messages en model latent dirichlet al location lda author topic model adding key attribute distribution recipient discovery topics ing relationships people 
#@A McCallum, A Corrada-Emmanuel, Xuerui Wang
#t2009
#c238
#index557
#id569
#%
#!

#*estimation prediction stochastic paper extend approach case relations set possible values number classes attributes class structure unobserved latent fi conditional model using generalization model mixture model classes vertices 
#@Krzysztof Nowicki, Tom A.B. Snijders
#t2009
#c133
#index558
#id570
#%
#!

#*group individual voting behavior year group cohesion voting similarity analysis group cohesion studies common analyses parliament published article provides group cohesion analysis using agreement index modified version index advanced voting similarity analyses new easy understand way results provided agreement index operates party group level voting similarity analyses able illustrate individual level article testing voting similarity methods provides insight recent voting behaviour 
#@A Pajala, A Jakulin, W Buntine
#t2009
#c19
#index559
#id571
#%
#!

#*hidden markov models new algorithms applications hidden markov models hmms applied important problems introduce new learning algorithm hmms unlike classical baum algorithm applied line batch usual likely path approximation left right hmms states trained represent protein including cases models 
#@Pierre Baldi, Yves Chauvin, Tim Hunkapiller, Marcella A. McClure
#t2009
#c190
#index560
#id572
#%
#!

#*statistical decision theory bayesian analysis book topic coverage references examples problem sets book excellent addition mathematical library mathematical new edition author added substantial material bayesian analysis including new important topics empirical hierarchical bayes analysis bayesian calculation bayesian communication group decision making changes book used self contained introduction bayesian analysis addition decision theoretic portion text updated including new modern topics multivariate estimation 
#@J. O. Berger
#t2009
#c30
#index561
#id573
#%
#!

#*protein modeling using hidden markov models analysis apply hidden markov models hmms statistical modeling multiple protein variant expectation maximization em algorithm known used obtain statistical model detailed series experiments taken sequences produced statistical model primary sequences using 
#@D Haussler, A Krogh, IS Mian, K Sjolander
#t2009
#c256
#index562
#id574
#%
#!

#*study grammatical inference grammatical inference inductive process discovering grammar language basis finite samples language study useful inference procedures formal basis procedures states general grammatical inference problem formal languages reviews previous work states position evaluation measures indicates solution particular class grammatical inference problems based assumed probabilistic structure 
#@James Jay Horning
#t2009
#c50
#index563
#id575
#%
#!

#*estimation stochastic context free grammars using inside outside algorithm combination transformation algorithm stochastic context free grammars inside outside algorithm allows define method estimation rule probabilities stochastic context free grammars time complexity inside outside algorithm transformation algorithm stochastic context free grammars characteristic grammar proper does single rules stochastic context free grammars normal form 
#@K. Lari, S. Young
#t2009
#c304
#index564
#id576
#%
#!

#*learning ordered examples connectionist learning models considerable empirical success hard characterize learn learning finite state languages example domain extensively studied provide help understand connectionist learning major problem traditional learning assumes storage examples connectionist principles paper presents provably correct algorithm minimum state deterministic finite state complete ordered sample using limited total storage storing example algorithm iterative strategy uses stage current encoding data considered far single sample crucial advantages algorithm total space used course learning encoding finite prefix sample polynomial size inferred minimum state deterministic algorithm relatively efficient time implemented connectionist version algorithm properties connectionist version requires structure usual models implemented using connectionist machine finite working storage identify arbitrary 
#@Sara Porat, Jerome A. Feldman
#t2009
#c23
#index565
#id577
#%
#!

#*introduction hidden markov models basic theory markov chains known close years past applied explicitly problems speech processing major reasons speech models based markov chains developed recently method optimizing parameters markov model match observed signal patterns method proposed applied speech processing research theory implementation markov modelling techniques greatly enhanced method leading wide range applications models purpose tutorial paper introduction theory markov models illustrate applied problems speech recognition 
#@Lawrence R. Rabiner, Bing-Hwang Juang
#t2009
#c305
#index566
#id578
#%
#!

#*prior integers estimation minimum description length earlier introduced estimation principle number required observed data extend classical maximum likelihood principle principle estimation number parameters statistical models addition values way parameters appear models model structures principle new way construct prior distribution integers makes sense parameter individual object real parameters integers precision prior determined prior integers optimizing precision 
#@Jorma Rissanen
#t2009
#c306
#index567
#id579
#%
#!

#*estimation inference compact coding systematic variation set data represented usual statistical model used data compact form possible considered random form parts states inferred estimates unknown parameters model second states data using optimal code based data probability distribution parameter estimates choosing model estimates compact coding leads interesting general inference procedure form great properties computationally approximate form developed relation methods explored 
#@C. S. Wallace, P. R. Freeman
#t2009
#c307
#index568
#id580
#%
#!

#*direct brain communication activity primary areas similar way real movement eeg based brain interfaces based classification eeg changes different types left hand right hand movement features power adaptive parameters extracted eeg areas array central areas classification features linear discrimination analysis neural networks used characteristic classifier set learning updated online feedback using procedure rapid result discrimination brain states right hand movement training time able operate eeg based control hand nearly classification accuracy specific 
#@G. Pfurtscheller, C. Neuper
#t2009
#c181
#index569
#id581
#%
#!

#*using neural networks eeg classification paper compares different neural networks used classify single eeg data interface short introduction time series classification given used classifiers described standard used standard method classification compared finite response use filters instead static weights allow temporal processing inside classifier theoretical comparison architectures presented results experiment different subjects given discussed results demonstrate higher performance compared standard 
#@Ernst Haselsteiner, G. Pfurtscheller
#t2009
#c308
#index570
#id582
#%
#!

#*eeg pattern states detection classification use eeg detect arousal states humans eeg patterns described features extracted time significant feature mean frequency used arousal states detection eeg mean frequency characterized hidden markov models hmms estimate possible arousal state hmm single current state transition probability considered global states local contextual effects global contextual effects hmms used estimate detect proposed model behavior measure correct rate subject study estimated values mean frequency hmm high correlation behavior measure high recognition rate makes arousal states detection furthermore hmms applicable artifacts real time arousal states detection given subject states turn states applications example used maintain long term arousal 
#@R. S. Huang, C. J. Kuo, L.-L. Tsai, O. T. C. Chen
#t2009
#c309
#index571
#id583
#%
#!

#*hmm used classification eeg data hidden markov models hmm introduced classification single eeg data brain interface hmms used classify parameters eeg data left right hand movement effects different types hmms recognition rate discussed furthermore comparison results achieved linear hmm presented 
#@B. Obermaier, Ch. Guger, G. Pfurtscheller
#t2009
#c310
#index572
#id584
#%
#!

#*gaussian observation hidden markov models eeg analysis using number synthetic data sets hmms detect levels correlation frequency typical eeg number hidden states chosen using cluster analysis derived experiments based use hmms gaussian observation trained multivariate extraction 
#@William D. Penny, Stephen J. Roberts
#t2009
#c83
#index573
#id585
#%
#!

#*coupled hidden markov models modeling interactive processes present methods hidden markov models hmms model systems multiple processes resulting models multiple state variables coupled matrices conditional probabilities introduce deterministic approximation maximum posterior map state estimation enables fast classification parameter estimation expectation maximization dynamic programming algorithm samples highest 
#@M. Brand, N. Oliver, A. Pentland 1997
#t2009
#c50
#index574
#id586
#%
#!

#*estimation coupled hidden markov models application interaction modelling coupled hidden markov models new tool model interactions state space observation space reveal classical tools correlation paper derive maximum likelihood equations parameters using expectation algorithm use models demonstrated data signal analysis 
#@Iead Rezek, Stephen J. Roberts
#t2009
#c311
#index575
#id587
#%
#!

#*input output hmms sequence processing consider problems sequence processing propose solution based discrete state model order represent past context introduce connectionist architecture having structure state model statistical interpretation input output hidden markov model trained estimation maximization em generalized em algorithms considering state missing data temporal credit assignment actual parameter estimation model presents similarities hidden markov models hmms allows map input sequences output sequences using processing style neural networks trained using learning paradigm hmms potentially taking advantage em algorithm demonstrate suited solving grammatical inference problems benchmark problem experimental results presented seven grammars showing adaptive models excellent generalization 
#@Yoshua Bengio, Paolo Frasconi
#t2009
#c284
#index576
#id588
#%
#!

#*hidden markov models present framework learning hidden markov models distributed framework derive learning algorithm based maximization em procedure maximum likelihood estimation standard baum update rules step algorithm solved combinatorial nature representation exact step intractable simple tractable mean derived empirical results set problems suggest mean approximation sampling alternatives expensive exact algorithm 
#@Zoubin Ghahramani, Michael I. Jordan
#t2009
#c23
#index577
#id589
#%
#!

#*mixed memory markov models complex stochastic processes mixtures simpler ones fi study markov models state spaces arise product random transition matrices models mixture simpler modelsfi parameters models interpretation expectation maximization em set generalized baum updates hidden markov models 
#@Lawrence K. Saul, Michael I. Jordan
#t2009
#c23
#index578
#id590
#%
#!

#*coupled hidden markov models complex recognition present algorithms training hidden markov models hmms model processes demonstrate conventional hmms vision task classifying actions hmms successful framework computing modeling classifying dynamic popular offer dynamic time training algorithm clear bayesian semantics framework makes strong assumptions generating signal single process having small number states extremely limited state memory single process model vision speech applications resulting low model performance coupled hmms provide efficient way problems offer superior training model robustness initial conditions 
#@M. Brand, N. Oliver, Alex Pentland
#t2009
#c217
#index579
#id591
#%
#!

#*modeling traffic coupled hmms consider problem modeling loop data collected data vector time series contains speed second sampling number assume measured speed generated hidden discrete variable represents free traffic point assume hidden variables 
#@Jaimyoung Kwon, Kevin Murphy
#t2009
#c312
#index580
#id592
#%
#!

#*event coupled hidden markov models time series data greatly taking account multiple cases speech corresponding video different time series interested coupled time series events coupled present algorithm used learning event coupled hidden markov results 
#@Trausti T. Kristjansson, Brendan J. Frey, T.Huang
#t2009
#c313
#index581
#id593
#%
#!

#*approximate learning dynamic models inference key component learning probabilistic models data learning temporal models requires complete potentially sequence furthermore data structures extremely large making process approximate inference algorithm processes prove bounds approximation 
#@Xavier Boyen, Daphne Koller
#t2009
#c190
#index582
#id594
#%
#!

#*algorithm approximate inference algorithm simple approximate inference algorithm dynamic bayesian networks similar fully version algorithm instead exact update step followed projection works distributions applied models exact update step intractable belief 
#@Kevin P. Murphy, Yair Weiss
#t2009
#c129
#index583
#id595
#%
#!

#*growth transformations functions paper look problem function defined primarily concerned case certain space polynomial coefficients defined results valid 
#@LE Baum, GR Sell
#t2009
#c314
#index584
#id596
#%
#!

#*maximum likelihood estimation multivariate mixture observations markov chains use probabilistic functions markov chain model certain speech signal extend estimation technique multivariate mixtures gaussian products mixtures problems original framework 
#@Bing-Hwang Juang, Stephen E. Levinson, M. Mohan Sondhi
#t2009
#c211
#index585
#id597
#%
#!

#*identifying comparative sentences text documents paper studies problem identifying comparative sentences text documents problem related quite different sentiment opinion sentence identification classification sentiment classification studies problem classifying document sentence based subjective opinion author important application area sentiment opinion identification business intelligence product manufacturer wants know consumers opinions products comparisons hand subjective objective furthermore comparison concerned object instead compares object example opinion sentence quality cd poor example comparative sentence quality cd good cd clearly sentences different information language quite different identifying comparative sentences useful practice direct comparisons ways evaluation important opinions individual object paper proposes study comparative sentence identification problem comparative sentences different types presents novel integrated pattern discovery supervised learning approach identifying comparative sentences text documents experiment results using types documents news articles consumer reviews products internet postings precision recall detailed results given paper 
#@Nitin Jindal, Bing Liu
#t2009
#c35
#index586
#id598
#%587,588,589,393,590,394,591,397,398,385,137,201,142,592,593,145,594,146,147,150,595
#!

#*mining sequential patterns given large database customer transactions transaction consists customer transaction time items transaction introduce problem mining sequential patterns databases present algorithms solve problem empirically evaluate performance using synthetic data proposed algorithms comparable performance performs little better 
#@Rakesh Agrawal, Ramakrishnan Srikant
#t2009
#c55
#index587
#id599
#%
#!

#*sequential pattern mining using representation introduce new algorithm mining sequential patterns algorithm especially efficient sequential patterns database long introduce novel depth search strategy integrates depth search space effective pruning mechanisms implementation search strategy combines representation database efficient support feature algorithm outputs new frequent online fashion experimental evaluation algorithm standard benchmark data literature algorithm outperforms previous work order magnitude 
#@Jay Ayres, Jason Flannick, Johannes Gehrke, Tomi Yiu
#t2009
#c3
#index588
#id600
#%
#!

#*simple rule based speech automatic speech tagging area natural language processing statistical techniques successful rule based methods paper present rule based speech automatically rules accuracy stochastic rule based advantages including reduction stored information required set meaningful rules 
#@Eric Brill
#t2009
#c315
#index589
#id601
#%
#!

#*wide coverage grammar english paper presents grammar development tool based tree grammar includes wide coverage syntactic grammar english various components discussed preliminary evaluation results parsing various corpora given results comparison ibm statistical parser natural language tool parser given introduction large project develop 
#@Christy Doran, Dania Egedi, Beth Ann Hockey, B. Srinivas, Martin Zaidel
#t2009
#c316
#index590
#id602
#%
#!

#*extracting knowledge evaluative text capturing knowledge free form evaluative texts entity challenging task new techniques feature extraction polarity determination strength evaluation proposed feature extraction particularly important task provides extracted knowledge work paper introduces improved method feature extraction existing unsupervised method including user specific prior knowledge evaluated entity turn task feature extraction term similarity mapping learned features user defined entity features results terms accuracy mapping reduction semantic redundancy features 
#@Giuseppe Carenini, Raymond T. Ng, Ed Zwart
#t2009
#c317
#index591
#id603
#%
#!

#*web data mining exploring contents usage data association rule mining important model data mining mining algorithms discover item associations rules data satisfy user specified minimum support minimum confidence constraints minimum number data cases rule cover predictive strength rule used database model assumes items data nature similar frequencies data case applications applications items appear frequently data appear set high rules involve rare items rules involve frequent rare items set low combinatorial frequent items associated possible ways called rare item problem paper proposes novel technique solve problem technique allows user specify multiple minimum supports items frequencies database rule mining different rules need satisfy different minimum supports depending items rules experiment results technique effective 
#@Olfa Nasraoui
#t2009
#c318
#index592
#id605
#%
#!

#*sentiment analysis capturing using natural language processing paper sentiment analysis approach extract sentiments associated positive negative specific subjects document instead classifying document positive negative essential issues sentiment analysis identify sentiments expressed texts expressions indicate positive negative opinions subject order improve accuracy sentiment analysis important identify semantic relationships sentiment expressions subject applying semantic analysis syntactic parser sentiment lexicon prototype achieved high precision depending data finding sentiments web pages news articles 
#@Tetsuya Nasukawa, Jeonghee Yi
#t2009
#c317
#index593
#id606
#%
#!

#*mining sequential patterns pattern growth approach sequential pattern mining important data mining problem broad applications difficult problem mining generate examine explosive number intermediate previously developed sequential pattern mining methods explore candidate generation test approach reduce number candidates examined approach efficient mining large sequence databases having numerous patterns long patterns paper propose projection based sequential pattern growth approach efficient mining sequential patterns approach sequence database projected set smaller projected databases sequential patterns projected database exploring frequent based initial study pattern growth based sequential pattern mining propose efficient method called offers ordered growth reduced projected databases improve performance technique developed comprehensive performance study shows cases outperforms based algorithm sequential pattern mining algorithm data integrated tested algorithms furthermore mining methodology extended mining sequential patterns user specified constraints high pattern growth approach lead extension efficient mining kinds frequent patterns frequent 
#@Jian Pei, Jiawei Han, Behzad Mortazavi-Asl, Jianyong Wang, Helen Pinto, Qiming Chen, Umeshwar Dayal, Meichun Hsu
#t2009
#c15
#index594
#id607
#%
#!

#*cross collection mixture model comparative text mining problem refer comparative text mining given set comparable text collections task comparative text mining discover latent common themes collections summarize similarity di collections common theme general problem interesting applications including business intelligence opinion summarization propose generative probabilistic mixture model comparative text 
#@ChengXiang Zhai, Atulya Velivelli, Bei Yu
#t2009
#c3
#index595
#id608
#%
#!

#*incorporate syntactic knowledge opinion mining development internet content web user generated content novel media content produced end users taken past years web especially proven numbers traditional tasks increase volume data users manual way automatic mining approaches opinion mining recent data mining technique information retrieval computational linguistics suitable information processing paper work main opinion mining topic extraction sentiment classification propose approaches issues respectively chinese based consideration syntactic knowledge blog data typical application evaluating data experiments results approaches tasks promising introduction future plans work paper intelligent 
#@Guang Qiu
#t2009
#c319
#index596
#id609
#%597,141,139,218,598,385,599,600,601,602,150,145,603,604,605
#!

#*rise user generated content web consumer end user browser internet connection content power product service developing required 
#@Frank Bell
#t2009
#c50
#index597
#id610
#%
#!

#*opinion holder extraction author authority opinion holder extraction research important opinions viewed different paper experience opinion analysis task focusing opinion holder extraction results english approach opinion holder extraction based discrimination author authority opinionated sentences evaluation results respect documents 
#@Yohei Seki
#t2009
#c35
#index598
#id611
#%
#!

#*topic sentiment mixture modeling facets opinions weblogs paper define problem topic sentiment analysis weblogs propose novel probabilistic model capture mixture topics sentiments simultaneously proposed topic sentiment mixture tsm model reveal latent topical facets weblog collection results ad hoc query associated sentiments provide general sentiment models applicable ad hoc topics specifically designed hmm structure sentiment models topic models estimated tsm extract topic life sentiment dynamics empirical experiments different weblog datasets approach effective modeling topic facets sentiments extracting dynamics weblog collections tsm model quite general applied text collections mixture topics sentiments potential applications search result summarization opinion tracking user behavior prediction 
#@Qiaozhu Mei, Xu Ling, Matthew Wondra, Hang Su, ChengXiang Zhai
#t2009
#c38
#index599
#id612
#%
#!

#*learning subjective corpora subjectivity tagging distinguishing sentences used present opinions evaluations sentences used present information numerous applications subjectivity tagging relevant including information extraction information retrieval paper identifies strong clues subjectivity using results method clustering words according distributional similarity lin small detailed manual annotation 
#@Janyce Wiebe
#t2009
#c103
#index600
#id613
#%
#!

#*automatic retrieval clustering similar words bootstrapping semantics text challenges natural language learning earlier research showed possible automatically identify words semantically similar given word based syntactic patterns words present approach step obtaining tree structure similar words different senses given word identified different type paper topic 
#@Dekang Lin
#t2009
#c320
#index601
#id614
#%
#!

#*measuring inference semantic orientation association evaluative word called semantic orientation positive semantic orientation indicates negative semantic orientation indicates semantic orientation direction positive negative degree strong automated measuring semantic orientation application text classification text filtering tracking opinions online analysis survey automated systems paper introduces method semantic orientation word statistical association set positive negative paradigm words instances approach evaluated based different statistical measures word association mutual information latent semantic analysis lsa method experimentally tested words including adjectives nouns manually labeled positive words negative words method accuracy test set accuracy algorithm classifying words 
#@Peter D. Turney, Michael L. Littman
#t2009
#c107
#index602
#id615
#%
#!

#*utility linguistic rules opinion mining online product reviews important opinion sources web paper studies problem determining semantic orientations positive negative opinions expressed product features reviews existing approaches use set opinion words purpose semantic orientations words context dependent paper propose use linguistic rules deal problem new opinion aggregation function extensive experiments rules function highly effective called opinion observer built 
#@Xiaowen Ding, Bing Liu
#t2009
#c35
#index603
#id616
#%
#!

#*efficient syntactic tagging corpora tree important linguistics requires large number sentences annotated syntactic information time consuming difficult consistency annotation manually paper presented new technique semi automatic tagging chinese text takes input chinese text outputs tagged sentence tree use grammar employ based reduce context dependent parser tagging mechanism works human machine cooperative way machine tagging rules human level improved step step rules annotation addition good consistency tagging guaranteed 
#@Ming Zhou Changning Huang
#t2009
#c19
#index604
#id618
#%
#!

#*parsing english link grammar develop formal grammatical called link grammar english grammar algorithms efficiently parsing link grammar power link grammars context free grammars encoding natural language grammars appears new written program general link parsing written link grammar english language performance 
#@Daniel Dominic Sleator, David Temperley
#t2009
#c107
#index605
#id619
#%
#!

#*novel word clustering algorithm based latent semantic analysis new approach proposed clustering words given vocabulary method based paradigm formulated context information retrieval called latent semantic analysis paradigm leads vector representation word suitable vector space clustering techniques applied distance measure selected space arises naturally problem formulation preliminary experiments indicate clusters produced intuitively clusters semantic nature approach prove useful conventional class based statistical language modeling techniques 
#@Bellegarda, J.R.Butzberger, Janyce Wiebe
#t2009
#c321
#index606
#id620
#%
#!

#*improving retrieval information external sources major successful retrieval external sources electronic databases tremendous variability words people use objects fact different authors use different words essentially idea means relevant objects fact word used refer different means objects retrieved statistical method called latent semantic indexing models implicit higher order structure association words objects improves retrieval performance additional large performance improvements achieved use differential term weighting iterative retrieval methods 
#@S. Dumais
#t2009
#c322
#index607
#id621
#%
#!

#*document space models using latent semantic analysis paper approach constructing mixture language models based notion semantics discussed end technique known latent semantic analysis lsa used approach corpus derived semantic information able model varying style text using information corpus texts clustered unsupervised manner mixture automatically created work previous work field information 
#@Yoshihiko Gotoh, Steve Renals
#t2009
#c323
#index608
#id622
#%
#!

#*improving predicting performance statistical language models sparse domains standard statistical language models gram models represent probability word sequences sparse data problems tasks large amounts domain specific text available focuses improving estimation domain dependent gram models using domain text data previous approaches estimating language models multi domain data characteristic variations style content domains 
#@Rukmini Iyer, Dr. Mari Ostendorf
#t2009
#c50
#index609
#id623
#%
#!

#*contextual spelling using latent semantic analysis contextual spelling errors defined use valid word particular sentence context traditional spelling words typically attempt identify words used sentence explore use latent semantic analysis correcting used words results compared earlier work based bayesian classifier 
#@Michael P. Jones, James H. Martin
#t2009
#c324
#index610
#id624
#%
#!

#*maximum entropy approach adaptive statistical language modeling statistical language model described successfully integrates long distance linguistic information knowledge sources existing statistical language models exploit immediate history text extract information documents history propose use pairs basic information elements allows model topic statistical evidence multiple sources combined traditionally linear variants used shown instead apply principle maximum entropy information source gives rise set constraints combined estimate constraints set probability functions consistent information sources function highest entropy set solution given consistent statistical evidence unique solution guaranteed exist iterative algorithm guaranteed converge framework extremely general phenomenon described terms statistics text incorporated adaptive language model based approach trained journal corpus showed perplexity reduction baseline speech reduced error rate feasibility incorporating diverse knowledge sources single unified statistical framework 
#@Ronald Rosenfeld
#t2009
#c325
#index611
#id625
#%
#!

#*index optimized query processing query processing important building block ranked retrieval applications ranging text data integration distributed aggregation network logs sensor data queries operate index lists conditions aggregate scores result candidates best implementation methods setting family threshold algorithms aim index early possible based lower bounds final scores result candidates procedure performs sequential disk accesses sorted index performing random accesses score uncertainty scheduling kinds accesses different index lists sequential accesses decision perform random accesses candidates prior literature studied scheduling issues access types current paper takes integrated view scheduling issues novel strategies outperform prior large main new principled scheduling methods based optimization sequential accesses cost model random accesses methods probabilistic estimators scores index list correlations performance experiments different datasets trec terabyte server logs methods achieved significant performance compared best previously known methods 
#@Holger Bast, Debapriyo Majumdar, Ralf Schenkel, Martin Theobald, Gerhard Weikum
#t2009
#c326
#index612
#id626
#%222,613,447,226,614,615,227,616,617,618,232,619,620,621,622,623,464,624,625,226,626,45,506,46,627,628,629
#!

#*adaptive query processing large shared databases resources exhibit widely characteristics assumptions time query submitted query processing result traditional static query optimization execution techniques environments paper introduce query processing mechanism called operators query plan runs characterize 
#@Ron Avnur, Joseph M. Hellerstein
#t2009
#c10
#index613
#id627
#%
#!

#*effect adding relevance information relevance feedback environment effects adding information relevant documents examined trec routing modified rocchio relevance feedback approach used varying number retrieved initial smart search varying number terms used initial recall precision evaluation query adding terms relevant documents increases does 
#@Chris Buckley, Gerard Salton, James Allan
#t2009
#c35
#index614
#id628
#%
#!

#*efficient query calculation distributed networks typically enable web media applications paper presents new algorithm answer queries objects highest values distributed existing algorithms threshold algorithm bandwidth number nodes propose new algorithm called phase uniform threshold fi reduces network bandwidth pruning away objects data paper presents sets results driven size network reduces network orders magnitude compared existing second proven instance optimal data series lower bound particular analysis shows using pruning parameter achieves network example ratio data series following 
#@Pei Cao, Zhe Wang
#t2009
#c327
#index615
#id629
#%
#!

#*minimal supporting expensive predicates queries paper addresses problem evaluating ranked queries expensive predicates major support expensive user defined predicates boolean queries believe support ranked queries important ranked queries need model user specific concepts preference relevance similarity dynamic user defined functions second middleware systems incorporate external predicates integrating autonomous sources typically accessible object queries fuzzy joins expensive essentially user defined operations dynamically multiple relations predicates dynamically defined rely index mechanisms provide zero time sorted output instead require object evaluate current standard sort framework ranked queries efficiently handle predicates completely objects sorting merging produce answers minimize expensive develop formal principle necessary determines required propose algorithm implementing principle provably optimal minimal cost scale easily experiments using real estate benchmark database synthetic datasets enables significant reduction orders magnitude faster standard scheme using complete 
#@Kevin Chen-Chuan Chang, Seung-won Hwang
#t2009
#c10
#index616
#id630
#%630,502,504,497,498,631,500,632,232,616,633,634,228,229,226
#!

#*optimizing selection queries multimedia repositories repositories multimedia objects having multiple types attributes image text increasingly common query attributes typically just set objects traditional relational query model filtering grade match associated object indicates object matches selection condition ranking furthermore unlike relational model users just ranked objects selection queries relatively small addition differences query model multimedia repositories allow access attributes object indexes paper investigate optimize processing selection queries multimedia repositories access characteristics repositories query model lead novel issues query optimization particular choice indexes used search strongly cost processing filtering condition define execution space search minimal set indexes minimal general problem optimal plan search minimal execution space np hard present efficient algorithm problem respect cost model execution space predicates query independent problem optimizing selection queries viewed cases evaluating traditional selection conditions problems viewed extended filtering problem techniques query processing optimization 
#@Surajit Chaudhuri, Luis Gravano, Amelie Marian
#t2009
#c15
#index617
#id631
#%
#!

#*combining fuzzy information overview assume object database scores 
#@Ronald Fagin
#t2009
#c328
#index618
#id632
#%
#!

#*efficient multi feature queries heterogeneous environments applications like multimedia databases management systems efficiently retrieving best matching collections data present new combine processing multi feature queries data stream combine different data distributions combining furthermore present retrieval strategy essentially speed 
#@Ulrich Guntzer, Wolf-Tilo Balke, Werner Kiesling
#t2009
#c329
#index619
#id633
#%
#!

#*supporting join queries relational databases ranking queries produce results ordered computed score typically queries involve joins users usually interested join results current relational query processors handle ranking queries efficiently especially joins involved paper address supporting join queries relational query processors introduce new rank join algorithm makes use individual orders inputs produce join results ordered scoring function idea rank join results join operation introduce physical query operators based variants ripple join implement algorithm operators non integrated pipelined execution plans address practical issues optimization heuristics integrate new join operators practical query processors implement new operators inside prototype database engine based experimental evaluation approach compares recent algorithms ranked inputs shows superior performance 
#@Ihab F. Ilyas, Walid G. Aref, Ahmed K. Elmagarmid
#t2009
#c330
#index620
#id634
#%225,226,227,228,616,635,500,232,451,619,636,637,638,633,506,639
#!

#*rank aware query optimization ranking important property needs fully supported current relational query recently rank join query operators proposed based rank aggregation rank join operators rank join results performing join new operators direct impact traditional query processing introduce rank aware query optimization framework fully integrates rank join operators relational query framework based extending dynamic programming algorithm define ranking interesting property generation rank aware query unlike traditional join operators optimizing rank join operators depends estimating input cardinality introduce probabilistic model estimating input cardinality cost rank join knowledge paper effort estimating needed input size optimal rank aggregation ranking plans challenging key integration rank join operators real world query processing experimentally evaluate framework query open source database management experiments framework accuracy proposed estimation 
#@Ihab F. Ilyas, Rahul Shah, Walid G. Aref, Jeffrey Scott Vitter, Ahmed K. Elmagarmid
#t2009
#c10
#index621
#id635
#%
#!

#*history histograms history histograms long rich detailed information step includes course histograms different scientific fields failures histograms information solutions given great variety related problems paper techniques entire history including future history currently given fixed space details events results highest biased limited set experiments semantic distance compressed form history relatively small 
#@Yannis Ioannidis
#t2009
#c331
#index622
#id636
#%
#!

#*integration structure indexes inverted lists methods proposed evaluate queries xml dbms queries specify path keyword constraints graph approaches optimized structures known structure indexes approaches based information retrieval style inverted lists propose strategy combines forms indexes query evaluation algorithm path expressions based strategy technique general applicable wide range choices structure indexes inverted list join algorithms experiments xml dbms integrating forms indexes consider issues evaluating path expression queries notion relevance ranking incorporated integrating techniques threshold algorithm proposed fagin et al obtain instance optimal algorithms computation 
#@Raghav Kaushik, Rajasekar Krishnamurthy, Jeffrey F. Naughton, Raghu Ramakrishnan
#t2009
#c55
#index623
#id637
#%
#!

#*query algebra optimization relational queries paper introduces provides principled framework support efficient evaluations queries relational database systems rdbms extending relational algebra query optimization previously query processing studied middleware fashion focusing specific outside core query engines contrast support ranking class database construct insight new ranking relationship viewed property data parallel membership property data model membership essentially rdbms support ranking clearly ad fundamental integration ranking rdbms membership boolean filtering supported extend relational algebra rank relational capture ranking property new support ranking class construct extended algebra present pipelined incremental model ranking query plans expressed based fundamental ranking principle queries propose dimensional algorithm extended plan space plans dimensions ranking membership propose based method estimate cardinality rank plans experiments framework accuracy proposed estimation model 
#@Chengkai Li, Kevin Chen-Chuan Chang, Ihab F. Ilyas, Sumin Song
#t2009
#c10
#index624
#id638
#%
#!

#*optimized query execution large search engines global page ordering large web search engines answer thousands queries second interactive response times major factor cost query given lengths inverted lists query terms increase size document collection range address issue ir database researchers proposed pruning techniques compute approximate term based ranking functions inverted lists years search engines incorporated new types ranking techniques exploit aspects structure web popularity page obtain improved results focus question techniques efficiently integrated query processing particular study pruning techniques query execution large engines case global ranking pages provided pagerank method addition standard term based approach pruning schemes case evaluate efficiency experimental search engine million web pages results significant potential techniques 
#@Xiaohui Long, Torsten Suel
#t2009
#c8
#index625
#id639
#%459,640,641,191,642,643,488,644,226,63,645,646,497,500,618,42,232,647,619,66,648,649,170,650,651,652,653,654,506,655,46,656,657,658,659,660,632,61,661
#!

#*adaptive processing queries xml ability compute matches xml queries importance increasing number large 
#@Amelie Marian, Sihem Amer-Yahia, Nick Koudas, Divesh Srivastava
#t2009
#c55
#index626
#id640
#%
#!

#*efficient query engine search paper presents novel engine efficient ranked retrieval xml documents data algorithm follows paradigm threshold algorithms query processing focus inexpensive sequential accesses index lists random accesses difficulties applying 
#@Martin Theobald, Ralf Schenkel, Gerhard Weikum
#t2009
#c8
#index627
#id641
#%
#!

#*query evaluation probabilistic guarantees queries based ranking elements multidimensional datasets fundamental building block kinds information best known general purpose algo evaluating queries fagin threshold algorithm ta fi user goal queries identify relevant novel data items use approximate variants ta reduce run time paper introduces family approximate algorithms based probabilistic index lists multidimensional data space order local scores various forms derived bounds employed predict high probability candidate items index precision efficiency developed methods experimentally evaluated based large web corpus structured data fi queries multidimensional datasets compute relevant interesting results partial match based similarity scores attribute values query conditions score tion function weighted building block information discovery arises important application classes web search engines scores based word occur statistics possibly combining criteria like text based relevance link based authority multimedia similarity search feature vectors images music video preference queries structured data product catalogs support data having major text fi best known general purpose method queries fagin threshold algorithm known ta proposed et et fi method attribute multidimensional data space index list access data items order local score 
#@Martin Theobald, Gerhard Weikum, Ralf Schenkel
#t2009
#c8
#index628
#id642
#%
#!

#*processing distributed queries ranking aware queries queries attention recently various contexts web multimedia retrieval relational databases distributed systems queries play critical role decision making related identifying interesting objects network load balancing paper study ranking aggregation problem distributed systems prior research problem data distributions account simply assuming uniform data distribution nodes real data sets general paper propose efficient algorithms consider data distributions different ways extensive experiments demonstrate advantages approaches terms bandwidth 
#@Hailing Yu, Hua-Gang Li, Ping Wu, Divyakant Agrawal, Amr El Abbadi
#t2009
#c332
#index629
#id643
#%
#!

#*framework combining preferences world wide web created available line information range potential choices time effort required sort propose formal framework combining user preferences address problem preferences used focus search queries order search results preference expressed user entity described set named fields field values certain type used match element type set preferences combined using generic combine operator value function providing great deal preferences combined way combination preferences yields preference providing property demonstrate power framework currently popular real life application special cases framework discuss implementation framework relational setting 
#@Rakesh Agrawal, Edward L. Wimmers
#t2009
#c10
#index630
#id644
#%
#!

#*efficient execution multi parametric ranked queries users need optimize selection objects weighting importance multiple object attributes optimization problems appear operations research applied mathematics life select home weighted function number attributes like distance price area capture queries definition preference queries use weight function relations attributes derive score database systems efficiently produce results preference query need evaluate weight function relation prefer answers preference queries efficiently using materialized views stored result preference query produced pipelined fashion using materialized view excellent performance given reasonable number materialized views provide algorithm selects number views given space constraints implemented algorithms proposed paper prototype called prefer operates commercial database management present results performance comparison comparing algorithms prior approaches using synthetic datasets results indicate proposed algorithms superior performance compared approaches materialized views execution time 
#@Vagelis Hristidis, Nick Koudas, Y. Papakonstantinou, Yannis Papakonstantinou, La Jolla Ca, La Jolla Ca
#t2009
#c333
#index631
#id645
#%
#!

#*using algorithm merging ranked results multimedia middleware distributed multimedia information allows applications access variety data different stored data sources specialized search capabilities user set objects ranked particular property combination properties fagin gives algorithm efficiently merging multiple ordered streams ranked results form new stream ordered combination ranks 
#@Edward Wimmers, Laura Haas, Mary Tork Roth, Christoph Braendli
#t2009
#c334
#index632
#id646
#%
#!

#*supporting incremental join queries ranked inputs paper problem multiple ranked data sets list arbitrary user defined input problem arises important applications dealing multiple ranked data sets use multimedia examples equally applicable traditional database optimal resource 
#@Apostol Natsev, Yuan-Chi Chang, J.R. Smith, Chung-Sheng Li, Jeffrey Scott Vitter
#t2009
#c8
#index633
#id647
#%
#!

#*sql paper study simple sql extension enables query explicitly limit cardinality query result examine impact query optimization run time execution components relational dbms presenting approach aggressive exploiting cardinality relational query plans results obtained empirical study conducted using db demonstrate sql extension illustrate approaches implementing 
#@Michael J. Carey, Donald Kossmann
#t2009
#c114
#index634
#id648
#%
#!

#*rank aggregation methods web consider problem combining ranking results context web main building meta search engines functions selecting documents based improving search precision word set techniques rank compare performance known primary goal work aggregation techniques 
#@Cynthia Dwork, Ravi Kumar, Moni Naor, D. Sivakumar
#t2009
#c38
#index635
#id649
#%
#!

#*ripple joins online aggregation present new family join algorithms called ripple joins online processing multi table aggregation queries database management dbms queries arise naturally interactive exploratory decision support applications traditional join algorithms designed minimize time query contrast ripple joins designed minimize time precise query result available measured length confidence ripple joins adaptive behavior processing properties data ripple joins user dynamically trade key performance factors line aggregation time updates run aggregate confidence length update ripple joins implemented existing dbms using overview methods used compute confidence optimize ripple join aspect ratio parameters experiments initial implementation algorithms dbms time required produce precise online estimates orders magnitude smaller time required best join algorithms pro exact answers 
#@Peter J. Haas, Joseph M. Hellerstein
#t2009
#c10
#index636
#id650
#%
#!

#*optimization parallel query execution plans authors approach optimization query execution plans multi user parallel database machine based shared memory multi processor disk array main difficulties optimization problem time unknown parameters available size number free processors search space possible parallel plans authors deal problems novel phase optimization strategy dramatically reduces search space allows run time parameters significantly plan present phase strategy experimental evidence indicate produces optimal plans 
#@Wei Hong, Michael Stonebraker
#t2009
#c335
#index637
#id651
#%
#!

#*ranked inputs practice ranked inputs essential requirement database applications ranking search results multiple search engines answering multi feature queries multimedia retrieval systems introduce new practical pipelined query operator produces global rank input ranked streams based score function output serve valid input operators query operator support 
#@Ihab F. Ilyas, Walid G. Aref, Ahmed K. Elmagarmid
#t2009
#c8
#index638
#id652
#%
#!

#*pipelined join operator wide area distribution significant performance problems traditional query processing techniques data access link load pipelined query execution promising approach environments allows scheduling adjust properties data developed non join operator called small memory 
#@Tolga Urhan University, Tolga Urhan, Michael J. Franklin
#t2009
#c336
#index639
#id653
#%
#!

#*searching web offer overview current web search engine design generic search examine engine component turn cover local web page storage indexing use link analysis boosting search performance common design techniques components presented draw literature experimental search engine emphasis fundamental concepts results performance analyses conducted 
#@Arvind Arasu, Junghoo Cho, Hector Garcia-Molina, Andreas Paepcke, Sriram Raghavan
#t2009
#c337
#index640
#id654
#%
#!

#*distributed query processing using partitioned inverted files paper study query processing distributed text database real distributed implementation offers query distributed network workstations model server document indexed inverted adopt strategies index partitioning distributed local index partitioning global index strategies documents ranked using vector space model document filtering fast evaluate compare impact index partitioning strategies query processing experimental results retrieval efficiency framework global index outperforms local index 
#@Claudine Santos Badue, Ricardo A. Baeza-Yates, Berthier A. Ribeiro-Neto, Nivio Ziviani
#t2009
#c36
#index641
#id655
#%
#!

#*finding link structures world wide web recently given fi usually combined content analysis sense study link 
#@Allan Borodin, Gareth O. Roberts, Jeffrey S. Rosenthal, Panayiotis Tsaparas
#t2009
#c38
#index642
#id656
#%
#!

#*scale services web network yahoo past years despite scale growth rates rapid evolution content features sites scale services like available major web sites similar availability requirements article look basic model services focusing key challenges face high availability evolution growth developing principles problems 
#@Eric A. Brewer
#t2009
#c338
#index643
#id657
#%
#!

#*documents given documents mathematical capture informal roughly roughly basic idea reduce issues set problems easily evaluated process random sampling furthermore evaluated using size sample paper discusses mathematical properties measures cient implementation sampling process using 
#@AZ Broder
#t2009
#c339
#index644
#id658
#%
#!

#*evaluating performance distributed architectures information retrieval using variety workloads information internet offers access increasing number document collections order users effectively access collections information retrieval ir systems provide distributed access article explore achieve scalable performance distributed collection sizes ranging implement fully functional distributed ir based version inquery simulation model measure performance function parameters rate number document collections query query term frequency number answers mixture results important model query document significantly performance based results recommend simple changes prototype evaluate changes using significant resource information retrieval difficult generate workloads resources architecture workloads demonstrate organizations response time increases performance scales number processors scalable architecture includes small number large number servers communicate 
#@Brendon Cahoon, Kathryn S. McKinley, Zhihong Lu
#t2009
#c37
#index645
#id659
#%
#!

#*automatic resource list analyzing structure associated text design evaluation arc automatically list authoritative web resources sufficiently broad topic goal arc resource lists similar provided yahoo fundamental difference services construct lists manually combination human automated effort arc operates fully automatically evaluation arc yahoo resource lists human users evaluation suggests resources arc frequently better lists resources manually compiled classified topic provide examples arc resource lists reader examine 
#@Soumen Chakrabarti, Byron Dom, Prabhakar Raghavan, S. Rajagopalan, D. Gibson, J. Keinberg
#t2009
#c340
#index646
#id660
#%
#!

#*optimizing queries image databases digital libraries image retrieval queries based similarity objects using feature attributes like shape texture color text multi feature queries return ranked result set instead exact matches user wants ranked objects present new algorithm called combine european combining multi feature result lists correct retrieval ranked results score aggregation combining function used including weighted queries compared algorithm developed improved termination condition combination heuristic control flow itself particular score distribution ranked results computed output dramatically improve performance particular non uniform score distributions practical data indicate efficiency factor data observed speed factors larger performance results scale different database sizes numbers result sets combine 
#@Ulrich Guntzer, Wolf-Tilo Balke, Werner Kiesling
#t2009
#c341
#index647
#id661
#%
#!

#*topic sensitive pagerank original pagerank algorithm improving search query results single pagerank vector computed using link structure web capture importance web pages independent yield accurate search results propose computing set pagerank vectors biased set representative topics capture notion importance respect particular using 
#@T. Haveliwala
#t2009
#c38
#index648
#id662
#%
#!

#*inverted file partitioning schemes multiple disk systems multiple disk systems disk approach data applications information retrieval partition files multiple exploit potential balanced distribution important good performance parallel information retrieval using inverted affected partitioning 
#@Byeong-Soo Jeong, Edward Omiecinski
#t2009
#c75
#index649
#id663
#%
#!

#*stochastic approach link structure analysis effect today searching information www usually query term based search engine engine return result list web sites contents matches topic queries searches result set contained link structure information pages linked used augment 
#@Ronny Lempel, Shlomo Moran
#t2009
#c46
#index650
#id664
#%
#!

#*optimizing result prefetching web search engines segmented indices study process search segmented indices serve queries investigate number search engines query processing phase search engine users observed pages results behavior suggests prefetching processing initial query cient results requested user policy prefetching retrieving just page search results use resources argue certain behavior users engines constant result pages query query processing model search en segmented indices analyze prefetching policies based costs determine constant prefetching results applicable inverted applicable processing short 
#@Ronny Lempel, Shlomo Moran
#t2009
#c337
#index651
#id665
#%
#!

#*scalable distributed architectures information retrieval information internet information retrieval ir systems scale provide scalable performance rapidly increasing data workloads critical design generation information retrieval systems studies scalable distributed ir architectures provide response maintain retrieval accuracy distributed architectures exploit information retrieval cluster parallel ir servers using symmetric use partial collection replication selection collection selection search small data maintaining retrieval accuracy investigate using partial collection replication ir systems examine query real systems select partial based relevance load balance original collection updating strategies results sufficient query partial replication information retrieval proposed selection algorithm effectively selects relevant partial inexpensive implement evidence indicates partial replication achieves better performance queries selection algorithm finds similarity non identical queries increases observed use perform detailed performance evaluation distributed ir architectures explore best build parallel ir servers using symmetric evaluate performance partial collection replication collection selection compare performance partial collection replication collection partitioning collection selection 
#@Zhihong Lu, Kathryn S. McKinley
#t2009
#c342
#index652
#id666
#%
#!

#*building distributed text index web identify crucial design issues building distributed inverted index large collection web pages introduce novel technique core index building substantially reduces index construction time propose storage scheme creating managing inverted files using embedded database suggest compare different strategies global statistics distributed inverted indexes finally present performance results experiments distributed web indexing implemented 
#@Sergey Melnik, Sriram Raghavan, Beverly Yang, Hector Garcia-Molina
#t2009
#c38
#index653
#id667
#%
#!

#*search yields high quality pages paper examines average page quality time pages web million unique use connectivity based metric pagerank measure quality web graph search order good strategy tends discover high quality pages early 
#@Marc Najork, Janet L. Wiener
#t2009
#c343
#index654
#id668
#%
#!

#*pagerank citation ranking order web importance subjective matter depends readers interests knowledge relative importance web pages paper describes pagerank method rating web pages effectively measuring human attention compare pagerank random efficiently compute pagerank large numbers pages apply pagerank search user navigation 
#@Lawrence Page, Sergey Brin, Rajeev Motwani, Terry Winograd
#t2009
#c344
#index655
#id669
#%
#!

#*intelligent probabilistic combination link content information pagerank pagerank algorithm used google search engine greatly improves results web search taking account link structure pagerank assigns page score number times random page page page following page equal propose improve page rank using intelligent probabilistic model relevance page efficient execution algorithm query time possible pre computing time queries experiments large subsets web indicate algorithm significantly outperforms pagerank quality pages efficient used today large search 
#@Matthew Richardson, Pedro Domingos
#t2009
#c190
#index656
#id670
#%
#!

#*search engines web dynamics paper study dimensions web dynamics context large scale growth update dynamics clearly represent challenges problems arise components reference search engine 
#@Knut Magne Risvik, Rolf Michelsen
#t2009
#c46
#index657
#id671
#%
#!

#*design implementation high performance distributed web broad web search engines specialized search tools rely large collections pages indexing web millions hosts period issues robustness major addition performance network resources taken account order achieve high reasonable 
#@Vladislav Shkapenyuk, Torsten Suel
#t2009
#c55
#index658
#id672
#%
#!

#*peer peer architecture scalable web search information retrieval consider problem building based search engine massive document collections prototype called open distributed search engine architecture currently development group provides highly distributed global indexing query execution service used content inside outside network different approaches search assumes search engine architecture global index structure distributed network overview proposed discuss basic design choices main focus efficient query execution discuss recent work queries database community applied highly distributed environment preliminary simulation results real search engine log terabyte web collection indicate good scalability approach 
#@Torsten Suel, Chandan Mathur, Jo-wen Wu, Jiangong Zhang, Alex Delis, Mehdi Kharrazi, Xiaohui Long, Kulesh Shanmugasundaram
#t2009
#c345
#index659
#id673
#%
#!

#*performance inverted indices distributed text document retrieval systems performance distributed text document retrieval systems strongly organization inverted paper compares performance impact various physical organizations inverted present new database simulation experiments determine variables influence response time leads set design trade range hardware 
#@Anthony Tomasic, Hector Garcia-Molina
#t2009
#c346
#index660
#id674
#%
#!

#*implementations partial document ranking using inverted files commercial text retrieval systems employ inverted files improve retrieval paper concerns implementations document ranking based inverted methods implementing weighting strategy term inverse document frequency basic idea heuristic process query terms order documents possible identified 
#@Wai Yee Peter Wong, Dik Lun Lee
#t2009
#c63
#index661
#id675
#%
#!

#*missing probabilistic model document content hypertext connectivity joint probabilistic model modeling contents inter connectivity document collections sets web pages research paper model based probabilistic factor decomposition allows identifying principal topics collection authoritative documents topics furthermore relationships topics order build predictive model link content applications approach information retrieval search topic identification query disambiguation focused web web analysis 
#@David Cohn, Thomas Hofmann
#t2009
#c347
#index662
#id676
#%
#!

#*image information retrieval overview current research paper provides overview current research image information retrieval provides outline areas future broad focuses aspects image research ir text based retrieval content based retrieval image information retrieval review image retrieval evaluation studies similar information science image retrieval 
#@A. Goodrum
#t2009
#c348
#index663
#id677
#%
#!

#*automatic image annotation retrieval using cross media relevance models libraries traditionally used manual image indexing retrieving image manual image annotation expensive procedure great automatic ways retrieve propose automatic retrieving images based training assume regions image small vocabulary 
#@Jiwoon Jeon, Victor Lavrenko, R. Manmatha
#t2009
#c35
#index664
#id678
#%
#!

#*introduction variational methods graphical models paper presents tutorial introduction use variational methods inference learning graphical models bayesian networks markov random fields present number examples graphical models including database belief network machine variants hidden markov models run exact inference algorithms introduce variational methods exploit large numbers original graphical model graphical model inference efficient inference model provides bounds probabilities original model general framework generating variational transformations based finally return examples demonstrate variational algorithms formulated case 
#@Michael I. Jordan, Zoubin Ghahramani, Tommi Jaakkola, Lawrence K. Saul
#t2009
#c23
#index665
#id679
#%
#!

#*model multimedia information retrieval research multimedia information retrieval recently feature research trend independent fields resulting methods systems long run result slow primary goal study integration methods techniques model unified perspective produced model offers retrieval media text images dimensions form content way similarity based methods semantics based ones providing design systems able provide generalized multimedia retrieval service existing forms retrieval combined desired model formulated terms fuzzy description logic role directly models semantics based retrieval offers ideal framework integration multimedia multidimensional aspects retrieval model accounts relevance feedback text image retrieval integrating known techniques taking account user implementation model addressed presenting decomposition technique reduces query evaluation processing simpler solved means widely known methods text image retrieval semantic prototype multidimensional image retrieval presented shows decomposition technique work significant 
#@Carlo Meghini, Fabrizio Sebastiani, Umberto Straccia
#t2009
#c112
#index666
#id680
#%
#!

#*probabilistic framework semantic video indexing filtering retrieval semantic filtering retrieval multimedia content crucial efficient use multimedia data repositories video query semantic keywords difficult problems multimedia data retrieval difficulty mapping low level video representation high level semantics multimedia content access problem multimedia pattern recognition problem propose probabilistic framework semantic video indexing support filtering retrieval facilitate efficient content based access map low level features high level semantics propose probabilistic multimedia objects examples movies include music semantic concepts videos model interaction explicitly propose network using probabilistic models site using bayesian belief network demonstrate application framework semantic indexing demonstrate detection performance significantly improved using relationships account heterogeneous features support detection based inference reasoning 
#@H.Naphide, T.Huang
#t2009
#c349
#index667
#id681
#%
#!

#*estimating dirichlet distribution dirichlet distribution compound variant dirichlet multinomial basic models data vocabulary words text document maximum likelihood estimate distributions available closed form paper describes simple efficient iterative schemes obtaining parameter estimates models case fixed point generalized provided 
#@Thomas P. Minka
#t2009
#c50
#index668
#id682
#%
#!

#*modeling predicting personal information dissemination behavior paper propose new way automatically model predict human behavior information analyzing content personal communications personal profile called established individual based novel algorithm incorporating content time information simultaneously used personal social management clusters provide view informal networks organization management new algorithm developed based combination dynamic algorithms social network field semantic content classification methods natural language processing machine learning tested email corpus report experimental results including filtering prediction recommendation capabilities personal behavior based models instance person specific email predicted ones personal social network content analysis experimental results prediction accuracy proposed adaptive algorithm better social network based predictions better model based latent dirichlet allocation social network online systems developed allow interactive exploration discussed 
#@Xiaodan Song, Ching-Yung Lin, Belle L. Tseng, Ming-Ting Sun
#t2009
#c3
#index669
#id683
#%670,671,672,673,674,675,676,655,170,677,678,556,181,176,360,432,679,680,224
#!

#*know know work information age discuss research personal social networks traditional resources resources networks social networks key sources information rapidly characterized fewer reliable resources personal social network fast alternative traditional transactions today 
#@BA Nardi, S Whittaker, H Schwarz
#t2009
#c350
#index670
#id684
#%
#!

#*structure article agreement approach organizational clusters individuals potentially aspects including structure network relations building recent work concerning defined embedded person research examines concerning relations results support idea relative general higher agreement concerning embedded organizations 
#@David Krackhardt, Martin Kilduff
#t2009
#c351
#index671
#id685
#%
#!

#*integrating communication information personal social networks short years massive use personal digital hybrid devices integrate internet services devices communicate traditional computers internet transformed computational environments communication spaces goal research integrate communication traditional information functions computational devices provide organizing principle advanced user interfaces information communication services single integrated based research communication patterns answer models personal social networks research showed people considerable effort maintaining links networks networks significant organizing principle work information article report study communication informed development software prototype user experiments 
#@BA Nardi, S Whittaker, Ellen Isaacs, Mike Creech, Jeff Johnson, John Hainsworth
#t2009
#c352
#index672
#id686
#%
#!

#*discovering shared interests people using graph analysis important problem users large networks discover resources data paper focus people particular interests usual approach group lists explicitly data assumes lists list present alternative approach support dynamically adaptive notion shared interests approach interests history mail communication using set heuristic graph algorithms demonstrate algorithms apply ing data collected sites using algorithms able lists people far data collection sites algorithms present powerful privacy propose believe use algorithms possible applications believe privacy including discovering resources people file data 
#@Michael F. Schwartz, David C. M.Wood
#t2009
#c352
#index673
#id687
#%
#!

#*web combining social networks collaborative filtering paper appears communications studies shown effective dissemination information organization network fi social network important organizational structure tasks immediate local 
#@Henry A. Kautz, Bart Selman, Mehul A. Shah
#t2009
#c7
#index674
#id688
#%
#!

#*self organization identification web communities despite nature web web self communities highly related pages efficiently identified based 
#@Gary William Flake, S Lawrence, C. Lee Giles, Frans Coetzee
#t2009
#c188
#index675
#id689
#%
#!

#*email automated discovery community structure organizations method automatic identification communities practice email logs use algorithm rapidly communities graph representing information apply algorithm email corpus nearly million messages collected method effective identifying true communities formal informal scale free approach enables identification studies evaluation results 
#@JR Tyler, Dennis M. Wilkinson, BA Huberman
#t2009
#c353
#index676
#id690
#%
#!

#*models network data chapter statistical methods network consider models represented result usually non observed small changes occurring observed 
#@Tom Snijders Ics, Tom A.B. Snijders
#t2009
#c354
#index677
#id691
#%
#!

#*link prediction problem social networks given social network infer new interactions likely occur near future question link prediction problem develop approaches link prediction based measures analyzing proximity experiments large networks suggest future interactions extracted network topology detecting 
#@David Liben-Nowell, Jon M. Kleinberg
#t2009
#c32
#index678
#id692
#%
#!

#*relational evolutionary expert modeling develop novel user modeling technology dynamically update person profile enterprise environment technology enhance collaboration finding experts training instead using traditional search methods keyword match propose use relational evolutionary graph models experts used mining retrieval visualization conduct experiments building researchers research paper collection experiments demonstrate mining matching efficiently achieved based proposed relational evolutionary graph models 
#@Xiaodan Song, Belle L. Tseng, Ching-Yung Lin, Ming-Ting Sun
#t2009
#c355
#index679
#id693
#%
#!

#*line new event detection tracking work discuss evaluate solutions text classification problems associated events reported line source news present solutions related classification problems new event detection event clustering event tracking primary focus new event detection goal identify news stories previously reported stream news radio present algorithm new event detection analyze effects incorporating domain properties classification algorithm explore solution models temporal relationship news stories investigate use proper noun phrase extraction capture contained news results new event detection suggest previous approaches document clustering provide good basis approach new event detection improvements classification accuracy obtained domain properties news modeled new event detection related problem event clustering goal group stories discuss event investigate line clustering approach new event detection evaluate existing cluster comparison strategies previously used document retrieval results suggest strategies produce different events line single link strategy extended model domain properties faster effective approaches explore test representation issues context event tracking classifier event formulated sample stories classifier used subsequent news stream documents related event 
#@James Allan, Ron Papka, Victor Lavrenko
#t2009
#c35
#index680
#id694
#%
#!

#*multiscale topic modeling evolution topics time great value automatic summarization analysis large document collections work propose new probabilistic graphical model address issue new model multiscale topic model employs non homogeneous poisson processes model generation word evolution topics modeled multi scale analysis using new features model modeling evolution topics various time scales resolution allowing user time scales experiments science data using new model interesting patterns topics new model comparable lda predicting unseen data demonstrated perplexity experiments 
#@Ramesh Nallapati, Susan Ditmore, J Lafferty, Kin Ung
#t2009
#c3
#index681
#id695
#%335,176,430,539,338,682,203,683,684,685,686,421,687
#!

#*probabilistic approach automatic keyword indexing previously published research words tend frequency distributions described single poisson distribution 
#@Stephen P Harter
#t2009
#c53
#index682
#id696
#%
#!

#*bayesian multiscale models poisson processes focus article new class models bayesian multi scale models phenomena characteristics observed data modeled discrete signal time series independent poisson underlying intensity function potentially structure multiple scales fundamental analysis data estimation underlying intensity function problem particular addressed knowledge intensity goal itself serve initial step prior higher level analyses detection classification 
#@Eric D. Kolaczyk
#t2009
#c133
#index683
#id697
#%
#!

#*multiscale hidden markov models bayesian image analysis bayesian multiscale image analysis powerful probabilistic graphs intuitively tractable addition natural useful framework modeling bayesian multiscale analysis computationally classical markov random field modelsfi probabilistic graph model called multiscale hidden 
#@Robert D. Nowak
#t2009
#c356
#index684
#id698
#%
#!

#*statistical multiscale framework poisson inverse problems paper describes statistical modeling analysis method linear inverse poisson data based novel multiscale framework itself multiscale analysis associated recursive partitioning underlying intensity multiscale factorization likelihood analysis prior probability distribution match factorization modeling underlying 
#@Robert D. Nowak, Eric D. Kolaczyk
#t2009
#c211
#index685
#id699
#%
#!

#*simple effective approximations poisson model probabilistic weighted retrieval poisson model term frequencies used suggest ways variables concerned document length frequency simple weighting tested trec test collection considerable performance improvements collection frequency weighting demonstrated 
#@S.E. Robertson, S. Walker
#t2009
#c35
#index686
#id700
#%
#!

#*topics time non markov continuous time model topical trends paper presents lda style topic model captures low dimensional structure data structure changes time unlike recent work markov assumptions discretization time topic associated continuous distribution generated document mixture distribution topics word occurrences documents meaning particular topic constant topics occurrence correlations change significantly time present results personal email years research papers years state addresses showing improved topics better prediction trends 
#@Xuerui Wang, A McCallum
#t2009
#c3
#index687
#id701
#%176,688,360,338,689,690,691,557,327,432,424,669,692,546
#!

#*mixed membership models scientific world scientific classification structure subjects topic labels submitted authors articles largely related traditionally established include broad field classifications physical sciences sciences social sciences classifications fields focusing sciences explore internal soft classification structure articles based semantic compare formal classifications model assumes fixed number internal categories characterized multinomial distributions words references soft classification article based article content category discuss model database features data relevant soft classification 
#@Elena Erosheva Stephen, Stephen Fienberg, J Lafferty
#t2009
#c357
#index688
#id702
#%
#!

#*hierarchical structure streams fundamental problem text data mining extract meaningful structure document streams time mail news articles natural examples streams characterized topics appear grow intensity period time away published literature particular research field seen exhibit similar phenomena longer time scale underlying text mining work area following topic document stream activity certain features frequency topic goal present work develop formal approach modeling way efficiently identified provide organizational framework analyzing underlying content approach based modeling stream using state appear naturally state ways viewed drawing models theory network traffic resulting algorithms highly efficient yield nested representation set hierarchical structure overall stream experiments mail research paper suggest resulting structures natural meaning terms content rise 
#@Jon M. Kleinberg
#t2009
#c113
#index689
#id703
#%
#!

#*generalized probability density function bounded random processes author developed probability density function pdf random processes bounded lower occurring bounds condition generalized pdf bounded probability density function db pdf developed methods application practical problems parameter estimation simulation random variables numerical example author developed probability density function pdf random processes bounded lower occurring bounds condition generalized pdf bounded probability density function db pdf developed methods application practical problems parameter estimation simulation random variables numerical example 
#@P.Kumaraswamy
#t2009
#c358
#index690
#id704
#%
#!

#*modeling word using dirichlet distribution multinomial distributions used model text documents capture phenomenon words document tend appear word appears likely appear paper propose dirichlet compound multinomial model dcm alternative multinomial dcm model additional degree allows capture experimentally dcm substantially better multinomial modeling text data measured perplexity using standard document collections dcm leads better classification multinomial model dcm performance comparable obtained multiple heuristic changes multinomial model 
#@Rasmus Elsborg Madsen, David Kauchak, Charles Elkan
#t2009
#c22
#index691
#id705
#%
#!

#*constructing statistical models word usage present automatically date tagged free text ranks groups semantic features based statistical use features related stories deal single 
#@Russell Swan, David Jensen
#t2009
#c359
#index692
#id706
#%
#!

#*opinion extraction semantic classification product reviews web contains wealth product reviews task opinion mining tool process set search results given item generating list product attributes quality features opinions poor mixed good begin identifying unique properties problem develop method automatically distinguishing positive negative reviews classifier information retrieval techniques feature extraction scoring results various metrics heuristics vary depending testing situation best methods work better traditional machine learning operating individual sentences collected web searches performance limited noise context complete web based tool simple method grouping sentences attributes results qualitatively quite useful 
#@K Dave, S Lawrence, DM Pennock
#t2009
#c281
#index693
#id707
#%694,695,395,696,397,398,601,697,316,698,699,700,145,416,701,702,703,704,705,409,706,707,708
#!

#*yahoo amazon sentiment parsing small web internet feasible continuous stream public sentiment world wide web quite issue consideration present methodology real time sentiment extraction domain web increase influence individuals stock market web based posting sentiment stock message important capture sentiment small index sentiment compiled paper technology extracting small sentiment web sources create index applications methodology make use natural language statistical algorithms automated classification messages web design classification algorithms different theoretical content view sentiment single posting message use multiple methods allows voting rules classification process enables elimination fuzzy messages better rule algorithms improves classification accuracy leads natural increase number messages classified fuzzy classifier achieves accuracy versus random classification accuracy compares human agreement message classification technology computationally efficient allowing access thousands messages applications evidence strong link market sentiment based approximately messages evidence sentiment based stock 
#@SR Das, MY Chen
#t2009
#c360
#index694
#id708
#%
#!

#*word grams analyze evaluate experimentally new probabilistic model word sequence prediction natural language based prediction trees using efficient data structures extend notion use bayesian approach based recursive possible efficiently maintain tree mixtures mixtures provably practically better performance single model evaluate model corpora low perplexity achieved relatively small mixture models suggests alternative practically widely used gram models 
#@Fernando C. N. Pereira, Yoram Singer, Naftali Tishby
#t2009
#c107
#index695
#id709
#%
#!

#*good smoothing performance statistically based techniques tasks spelling sense disambiguation improved estimate probability object seen good methods means estimating probabilities previously unseen objects use good methods requires smoothing step regions different accuracy difficult use 
#@William Gale
#t2009
#c361
#index696
#id710
#%
#!

#*model textual affect using real world knowledge paper presents novel way assessing natural language use previous approaches textual affect employed keyword lexical statistical methods hand models paper demonstrates new approach using large scale real world knowledge nature situations classify sentences basic categories 
#@Hugo Liu, Henry Lieberman, Ted Selker
#t2009
#c362
#index697
#id711
#%
#!

#*feature subset selection text learning paper describes known new methods feature subset selection large text data experimental comparison given real world data collected web users shows characteristics problem domain machine learning algorithm considered feature scoring measure selected problem domain consists given form small documents represented word vectors learning experiments naive bayesian classifier used 
#@Dunja Mladenic
#t2009
#c363
#index698
#id712
#%
#!

#*book using text categorization extracted information content based recommender systems suggest documents items services users based learning profile user examples containing information given items text categorization methods useful task generally rely unstructured text developed utilizes semi structured information items web using simple information extraction techniques initial experimental results demonstrate 
#@Raymond Mooney, Paul N. Bennett, Loriene Roy
#t2009
#c364
#index699
#id713
#%
#!

#*mining product web products important marketing customer relationship management costly collect analyze survey data manually paper presents new framework mining product internet automatically people opinions target products web pages uses text mining techniques obtain products advance generate basis human tested 
#@Satoshi Morinaga, Kenji Yamanishi, Kenji Tateishi, Toshikazu Fukushima
#t2009
#c365
#index700
#id714
#%
#!

#*learning surface text patterns question answering paper explore power surface text patterns open domain question answering systems order obtain optimal set patterns developed method learning patterns automatically tagged corpus built internet bootstrapping process providing hand examples question type patterns automatically extracted documents precision pattern average precision question type patterns applied answers new questions using trec question set report results cases answers determined trec corpus web 
#@Deepak Ravichandran, Eduard H. Hovy
#t2009
#c104
#index701
#id715
#%
#!

#*automatically generating extraction patterns text corpus based natural language processing systems rely text corpora manually annotated syntactic semantic particular previous dictionary construction systems information extraction used annotated training corpus form annotated input developed called extraction patterns using text based generated extraction 
#@Ellen Riloff
#t2009
#c366
#index702
#id716
#%
#!

#*affect analysis text using fuzzy semantic typing propose novel convenient natural language processing fuzzy logic techniques analyzing affect content free text main fast analysis visualization affect content decision making primary linguistic resource fuzzy semantic typing fuzzy affect lexicon important resources generated fuzzy affect category groups free text tagged affect categories lexicon affect 
#@Pero Subasic, Alison Huettner
#t2009
#c367
#index703
#id717
#%
#!

#*sharing recommendations finding relevant high quality information wide web difficult problem people know experimental addresses problem collaborative filtering approach works automatically recognizing recommendations web resources news messages 
#@Loren G. Terveen, William C. Hill, Brian Amento, David W. McDonald, Josh Creter
#t2009
#c7
#index704
#id718
#%
#!

#*unsupervised learning semantic orientation word corpus evaluative word called semantic orientation positive semantic orientation negative semantic orientation paper introduces simple algorithm unsupervised learning semantic orientation extremely large corpora method involves queries web search engine using mutual information results algorithm empirically evaluated using training corpus approximately words subset web indexed chosen search engine tested words positive negative algorithm accuracy test words include adjectives nouns accuracy comparable results achieved using complex stage supervised learning algorithm restricted determining semantic orientation adjectives 
#@Peter D. Turney, Michael L. Littman
#t2009
#c107
#index705
#id719
#%
#!

#*corpus study evaluative language paper presents corpus study evaluative language knowledge language useful applications text categorization summarization analyses agreement characteristics subjective language performed study yields knowledge needed design effective machine learning systems identifying subjective language 
#@Janyce Wiebe, Rebecca Bruce, Matthew Bell, Melanie Martin, Theresa Wilson
#t2009
#c368
#index706
#id720
#%
#!

#*identifying recognizing opinions subjectivity natural language refers aspects language used express opinions evaluations 
#@Janyce Wiebe, Theresa Wilson, Matthew Bell
#t2009
#c369
#index707
#id721
#%
#!

#*using compute term frequency document frequency corpus commonly used statistical natural language processing paper techniques working longer grams introduced compute frequency location gram sequence corpus length compute frequencies corpus number equivalence classes way computation reduced computation classes paper presents algorithms code used compute term frequency document frequency grams large corpora english corpus million words journal corpus million second paper uses frequencies interesting interested grams high mutual information mi joint term frequency higher expected assuming parts gram combine inverse document frequency compares document frequency model terms particular term frequency distributed collection mi tends phrases semantics independence assumption tends technical names good keywords information retrieval tend exhibit distributions documents combination mi better itself word extraction task 
#@Mikio Yamamoto, Kenneth Ward Church
#t2009
#c279
#index708
#id722
#%
#!

#*opinion observer analyzing comparing opinions web excellent source consumer opinions numerous web sites containing opinions customer reviews products discussion groups blogs paper focuses online customer reviews products makes proposes novel framework analyzing comparing consumer opinions products prototype called opinion observer implemented single visualization user able clearly product consumers terms various product features comparison useful potential customers product potential customer visual feature feature comparison consumer opinions products helps decide product buy product manufacturer comparison enables easily gather marketing intelligence product information second new technique based language pattern mining proposed extract product features pros cons particular type reviews features form basis comparison experimental results technique highly effective outperform existing methods significantly 
#@Bing Liu, Minqing Hu, Junsheng Cheng
#t2009
#c281
#index709
#id723
#%710,711,393,712,394,713,397,398,385,714,404,715,593,716,145,147,717,150,718
#!

#*mining using networks arising social behavior recent advances information retrieval hyperlinked corpora demonstrated links noisy information text investigate feasibility applying link based methods new applications domains specific application consider partition authors given topic context typical posting consists posting followed opinion author social behavior gives rise network vertices individuals links represent relationships interesting characteristic people frequently message behavior contrast www link graph linkage agreement common analyzing graph structure able effectively classify people contrast methods based statistical analysis text yield low accuracy datasets vocabulary used tends largely identical postings relatively words text 
#@Rakesh Agrawal, S. Rajagopalan, Ramakrishnan Srikant, Yirong Xu
#t2009
#c38
#index710
#id724
#%
#!

#*information extraction relational markov networks information extraction systems treat separate potential extractions independent cases considering different potential extractions improve overall accuracy statistical methods based graphical models conditional random fields shown effective approach learning accurate systems present new method employs relational markov networks generalization represent arbitrary dependencies extractions allows information extraction mutual influence possible extractions experiments learning extract protein names text demonstrate advantages approach 
#@Razvan C. Bunescu, Raymond J. Mooney
#t2009
#c104
#index711
#id725
#%
#!

#*web scale information extraction preliminary results manually querying search engines order large information error process search engines retrieve rank potentially human extract information multiple documents aims process large collections web autonomous domain independent scalable manner paper describes preliminary experiments instance running single machine able automatically extract probability fact trade precision recall paper architecture reports learned design large scale information extraction systems 
#@Oren Etzioni, Michael Cafarella, Doug Downey, Stanley Kok, Ana-Maria Popescu, Tal Shaked, Stephen Soderland, Daniel S. Weld, Alexander Yates
#t2009
#c370
#index712
#id726
#%
#!

#*information extraction hmm structures learned stochastic optimization recent research demonstrated strong performance hidden markov models applied information extraction task database corresponding phrases text documents problem selection state transition structure model paper demonstrates extraction accuracy strongly depends selection structure presents algorithm automatically finding good structures stochastic optimization 
#@Dayne Freitag, A McCallum
#t2009
#c175
#index713
#id727
#%
#!

#*conditional random fields probabilistic models labeling sequence data present conditional random fields framework building probabilistic models segment sequence data conditional random fields offer advantages hidden markov models stochastic grammars tasks including ability strong independence assumptions models conditional random fields avoid fundamental limitation maximum entropy markov models markov models based directed graphical models biased states states present iterative parameter estimation algorithms conditional random fields compare performance resulting models hmms synthetic natural language data 
#@J Lafferty, A McCallum, Fernando Pereira
#t2009
#c371
#index714
#id728
#%
#!

#*hierarchical approach wrapper induction tremendous information available web basis ability quickly develop information agents crucial problem component web based information agent set extract relevant data information sources novel approach wrapper induction based idea hierarchical information extraction hard problem extracting data complex document series extraction tasks introduce inductive algorithm high accuracy extraction rules based user labeled training examples labeling training data represents major using wrapper induction techniques experimental results does significantly better approaches hand requires orders magnitude fewer examples algorithms hand handle information sources existing techniques 
#@Ion Muslea, Steven Minton, Craig A. Knoblock
#t2009
#c372
#index715
#id730
#%
#!

#*robust metric opinion exploring affect text paper describes automated detecting polar expressions topic components approach nlp polar language extraction machine learning based topic classifier components making simple accurate assumption topical sentence contains polar language polar language topic subject matter evaluate components assumption corpus online consumer messages based components discuss measure overall sentiment particular topic expressed online messages different people propose use bayesian statistics form aggregate opinion metric metric introduced polarity topic facilitate statistically valid comparisons opinion multiple topics 
#@K Nigam, Matthew Hurst
#t2009
#c146
#index716
#id731
#%
#!

#*classifying semantic relations text crucial step goal automatic extraction information natural language text identification semantic relations sentences examine problem distinguishing seven relation types occur entities text problem identifying entities compare generative graphical models neural network using lexical syntactic semantic features finding help achieve high classification accuracy 
#@Barbara Rosario, Marti A. Hearst
#t2009
#c373
#index717
#id732
#%
#!

#*web data extraction based partial tree paper studies problem extracting data web page contains structured data records objective segment data records extract data items fields data database table problem studied researchers existing methods limitations class methods based machine learning requires human labeling examples web site interested extracting data process time consuming large number sites pages web second class algorithms based automatic pattern discovery methods make assumptions paper proposes new method perform task automatically consists steps identifying individual data records page extracting data items identified data records step propose method based visual information segment data records accurate existing methods step propose novel partial technique based tree matching partial means data fields pair data records aligned make data fields approach enables accurate multiple data records experimental results using large number web pages diverse domains proposed step technique able segment data records extract data accurately 
#@Yanhong Zhai, Bing Liu
#t2009
#c38
#index718
#id733
#%
#!

#*opinion retrieval blogs opinion retrieval document retrieval process requires documents retrieved ranked according opinions query topic relevant document satisfy criteria relevant query topic contains opinions query matter positive negative paper opinion retrieval algorithm traditional information retrieval ir component topic relevant documents document set opinion classification component documents having opinions results ir step component rank documents based relevance query having opinions query implemented algorithm working tested using trec blog track data automatic title runs result showed improvements map score best automatic runs track result higher state art opinion retrieval tested data set 
#@Wei Zhang, Clement T. Yu, Weiyi Meng
#t2009
#c32
#index719
#id734
#%393,720,397,385,586,721,142,722,723,145,144,146,724,725,726,727,728,729,730,731,732
#!

#*determining semantic orientation terms analysis sentiment classification recent text classification concerned topic document opinion rich set applications ranging tracking users opinions products political candidates expressed online customer relationship management functional extraction opinions text determination orientation subjective terms contained text determination term opinionated content positive negative paper present new method determining orientation subjective terms method based quantitative analysis terms terms given line use resulting term representations semi supervised term classification method present outperforms known methods tested recognized standard task 
#@Andrea Esuli, Fabrizio Sebastiani
#t2009
#c374
#index720
#id735
#%
#!

#*major topic detection application opinion summarization specific information sources summarizing discovered opinions important improve services improve products queries detecting opinions similar task topic detection sentence level opinions positive negative identifying events correlated opinions important paper proposes major topic detection mechanism capture main concepts embedded relevant document set opinion summarization relevant sentences related major topic document set determines opinion polarity relevant sentence finally positive negative sentences respectively 
#@Lun-Wei Ku, Li-Ying Lee, Tung-Ho Wu, Hsin-Hsi Chen
#t2009
#c35
#index721
#id736
#%
#!

#*effective approach document retrieval wordnet recognizing phrases noun phrases queries identified classified types proper names dictionary phrases simple phrases complex phrases document phrase content words phrase certain size sizes different types phrases different determined using decision tree phrases important individual terms documents response query ranked matching phrases given higher utilize wordnet word senses query terms sense query term determined words definition compound words considered possible query experimental results approach yields improvements best known results trec collections short title queries using web data 
#@Shuang Liu, Fang Liu, Clement T. Yu, Weiyi Meng
#t2009
#c35
#index722
#id737
#%
#!

#*overview trec blog track rise internet creation journal like web page logs created highly dynamic subset world wide web real world events blogs weblogs recently new medium called collection blogs internet new interesting research areas blogs interesting features entries added order high volume addition blogs created authors intended mechanism self expression extremely accessible blog software wide ranging blogs opinions traditional media tends focus blogs technology different genres blogs written specific topic personal life blog track year aim explore information behaviour purpose new large scale test collection trec blog collection created run track tasks main task opinion retrieval open task opinion retrieval task focuses specific aspect blogs opinionated nature blogs second task introduced allow influence determination suitable second task aspects blogs temporal event related nature blogs spam 
#@Iadh Ounis, Maarten de Rijke, Craig Macdonald, G Mishne, Ian Soboroff
#t2009
#c375
#index723
#id738
#%
#!

#*using appraisal groups sentiment analysis little work date sentiment analysis classifying texts positive negative orientation use semantic features used classification present new method sentiment classification based extracting analyzing appraisal groups good appraisal group represented set attribute values task independent semantic based appraisal theory semi automated methods used build lexicon adjectives classify movie reviews using features based combined standard bag words features report state art accuracy addition types appraisal appear significant sentiment classification 
#@Casey Whitelaw, Navendu Garg, Shlomo Argamon
#t2009
#c32
#index724
#id739
#%733,589,734,735,393,397,736,737,593,189,145,738,739,740
#!

#*learning subjective language subjectivity natural language refers aspects language used express opinions evaluations numerous natural language processing applications subjectivity analysis relevant including information extraction text categorization goal work learning subjective language corpora clues subjectivity generated tested including low frequency words adjectives identified using distributional similarity features examined working features generated different data sets using different procedures exhibit consistency performance better data sets addition article shows density subjectivity clues context strongly likely word subjective provides results annotation study assessing subjectivity sentences high density features finally clues used perform opinion recognition type text categorization genre detection demonstrate utility knowledge acquired article 
#@Janyce Wiebe, Theresa Wilson, Rebecca Bruce, Matthew Bell, Melanie Martin
#t2009
#c279
#index725
#id740
#%
#!

#*comparative study feature selection text categorization paper comparative study feature selection methods statistical learning text categorization focus aggressive dimensionality reduction methods evaluated including term selection based document frequency information gain mutual information mi test term strength effective experiments using neighbor classifier corpus 
#@Yiming Yang, Jan O. Pedersen
#t2009
#c22
#index726
#id741
#%
#!

#*multiple ranking strategies opinion retrieval blogs opinion retrieval task trec approach identifying opinions blog scoring posts separately various aspects associated expression opinion topic including sentiment analysis spam detection link based authority estimation separate approaches combined single ranking significant improvement content baseline 
#@G Mishne
#t2009
#c50
#index727
#id742
#%
#!

#*using blog properties improve retrieval paper describes simple heuristics improve opinion retrieval effectiveness using blog specific properties blog used increase retrieval scores blog posts published near time significant event related query inexpensive approach estimation used identify level opinion expressed query specific weights used change importance spam filtering different types queries overall methods combined non retrieval approaches result substantial improvements state art 
#@G Mishne
#t2009
#c376
#index728
#id743
#%
#!

#*trec blog track web information discovery integrated tool university library information science blog tracks opinion task trec goal opinion task public sentiment given entity target involves retrieving relevant blogs identifying contain opinions target matter blog test collection contains considerable noise blogs non english content non blog content text retrieval systems based hypothesis noise reduction non english blogs text improve topic opinion retrieval explored various noise reduction approaches effectively noise blog data valid content creating separate indexes noise assess noise reduction effect opinion blog retrieval task sequential topic retrieval followed opinion classification opinion retrieval approach apply traditional ir methods retrieve topic blogs ranks opinionated blogs based opinion scores generated opinion methods opinion module consists opinion term module identify opinions based frequency opinion terms terms occur frequently opinion blogs rare term module uses rare terms good opinion classification module uses module uses computational linguistics distribution similarity approach learn subjective language training data 
#@Kiduk Yang, Ning Yu, Alejandro Valerio, Hui Zhang
#t2009
#c52
#index729
#id744
#%
#!

#*trec blog enterprise tracks trec university blog track expert search task enterprise track complex interactive question answering task question answering track track paper reports results 
#@Douglas Oard, Tamer Elsayed, Jianqiang Wang, Yejun Wu, Pengyi Zhang, Eileen Abels, Jimmy Lin, Dagbert Soergel
#t2009
#c52
#index730
#id745
#%
#!

#*recognition classification noun phrases queries effective retrieval shown using phrases document retrieval leads higher retrieval effectiveness paper define types noun phrases present algorithm recognizing phrases queries existing tools combined phrase recognition algorithm tested using set web queries query log set trec queries experimental results algorithm yields high phrase recognition accuracy use baseline noun phrase recognition algorithm phrases trec queries document retrieval experiment conducted using trec queries phrases phrases recognized baseline noun phrase recognition algorithm phrases recognized algorithm respectively retrieval effectiveness better better demonstrates phrases queries does improve retrieval effectiveness better noun phrase recognition yields higher retrieval performance 
#@Wei Zhang, Shuang Liu, Clement T. Yu, Chaojing Sun, Fang Liu, Weiyi Meng
#t2009
#c32
#index731
#id746
#%
#!

#*query expansion using local global document analysis automatic query expansion long suggested technique dealing fundamental issue word information retrieval number approaches studied recently attention focused techniques analyze corpus discover word relationship global techniques analyze documents retrieved initial local feedback paper compare effectiveness approaches global analysis advantages local generally effective using global analysis techniques word phrase structure local documents produces results effective simple local feedback 
#@Jinxi Xu, W. Bruce Croft
#t2009
#c35
#index732
#id747
#%
#!

#*genre classification use textual features functional linguistic theory genre based text categorization developed feature sets representing different types indicate different genres structure texts express text using features enables analysis large scale rhetorical differences genres examining features important classification specific domain studied scientific articles historical experimental sciences physical respectively smo learning algorithm feature set achieved accuracy classifying articles according field field specific terms used features highly weighted features consistent differences historical experimental sciences empirical evidence notion multiple scientific methods 
#@Shlomo Argamon, Jeff Dodick
#t2009
#c377
#index733
#id748
#%
#!

#*introduction support vector machines comprehensive introduction support vector machines svms new generation learning based recent advances statistical learning theory svms state art performance real world applications text hand written recognition image classification analysis established standard tools machine learning data mining students book accessible material required good theory applications concepts introduced accessible self contained relevant literature web sites containing software ensure forms ideal starting point study equally book associated web site guide updated literature new applications line software 
#@Nello Cristianini, John Shawe-Taylor
#t2009
#c30
#index734
#id749
#%
#!

#*methods automatic extraction methods state art nlp tasks recent work suggests true large training corpora available results limited evaluation task individual classifiers work explores complex task automatic extraction million words examine results terms constraints complexity different contextual representations noise behaviour nlp systems large corpora 
#@James R. Curran
#t2009
#c378
#index735
#id750
#%
#!

#*rhetorical parsing texts surface based approach texts just simple sequences sentences complex artifacts highly rhetorical structure paper explores formed rhetorical structures automatically derived means surface form based algorithms algorithms identify phrases sentences rhetorical relations textual produce valid rhetorical structure trees natural language texts algorithms empirically corpus analysis phrases rely order rhetorical structure trees algorithms evaluated evaluation automatically manually constructed rhetorical structure trees evaluation shows automatically derived rhetorical structures successfully context text summarization 
#@Daniel Marcu
#t2009
#c279
#index736
#id751
#%
#!

#*sentiment analysis using support vector machines diverse information sources paper introduces approach sentiment analysis uses support vector machines svms diverse sources potentially information including measures phrases adjectives available knowledge topic text models using features introduced combined models shown effective past et al versions models experiments movie review data internet movie database demonstrate hybrid svms combine style feature based svms based real measures obtain superior performance producing best results published using data experiments using feature set topic information smaller dataset music reviews hand annotated topic reported results suggest incorporating topic information models yield improvement 
#@Mullen Tony, Collier Nigel
#t2009
#c379
#index737
#id752
#%
#!

#*fast training support vector machines using sequential minimal optimization chapter describes new algorithm training support vector machines sequential minimal optimization smo training support vector machine svm requires solution large quadratic programming optimization problem smo large problem series possible problems small problems solved using time consuming numerical optimization loop memory required smo linear training set size allows smo handle large training sets large matrix computation smo scales linear quadratic training set size various test problems standard projected conjugate gradient algorithm scales linear training set size smo computation time svm evaluation smo linear svms sparse data sets database smo fast database linear svms smo times faster algorithm 
#@John C. Platt
#t2009
#c380
#index738
#id753
#%
#!

#*analyzing appraisal automatically present method classifying texts automatically based subjective content apply method semantic orientation parts text believe content apply appraisal help distinguishing different types subjective texts movie reviews consumer product reviews 
#@Maite Taboada, Jack Grieve
#t2009
#c381
#index739
#id754
#%
#!

#*identifying distance using features chapter uses functional linguistic theory basis extracting semantic features documents focus determination role constructing distance using hierarchical model represents authors language choices possible construct feature representation superior computational efficiency usual bag words approach experiments context classification features create clear separation different distance approach aspects affect functional linguistic theory 
#@Maria Herke-Couchman, Casey Whitelaw, Jon Patrick
#t2009
#c103
#index740
#id755
#%
#!

#*opinion spam analysis evaluative texts web valuable source opinions products services events individuals recently researchers studied opinion sources product reviews posts blogs existing research focused classification summarization opinions using natural language processing data mining techniques important issue far opinion spam online opinions paper study issue context product reviews opinion rich widely used consumers product past years aggregate opinions product reviews high time study spam reviews best knowledge published study topic web spam email spam investigated extensively opinion spam quite different web spam email spam requires different detection techniques based analysis million reviews million amazon com opinion spam reviews widespread paper spam presents novel techniques detect 
#@Nitin Jindal, Bing Liu
#t2009
#c102
#index741
#id756
#%644,393,742,385,743,145,146,744,745,746,747
#!

#*web spam web refers actions intended search engines ranking pages higher recently web spam increased dramatically leading search results paper presents comprehensive current techniques believe help developing appropriate 
#@Zoltan Gyongyi, Hector Garcia-Molina
#t2009
#c144
#index742
#id757
#%
#!

#*detecting spam web pages content analysis paper web spam created pages web order influence results search engines traffic certain pages profit paper previously techniques automatically detecting spam pages examines effectiveness techniques using classification algorithms combined heuristics identify spam pages collection pages spam non spam pages 
#@Alexandros Ntoulas, Marc Najork, Mark Manasse, Dennis Fetterly
#t2009
#c38
#index743
#id758
#%
#!

#*spam web use search engine optimization techniques spam links search results paper focus type spam spam identify spam pages party domains pages traffic propose model describing end end spam present methodology analyzing identify domains layer using sets commercial keywords methodology useful search engines ranking algorithms spam spam pages identify serve spam pages 
#@Yi-Min Wang, Ming Ma, Yuan Niu, Hao Chen
#t2009
#c38
#index744
#id759
#%
#!

#*identifying link spam pages increasing importance search today web traffic effort create search engine spam link analysis important factors current commercial search engines ranking systems new kinds spam links building link technique link based ranking algorithms paper present algorithms detecting link automatically generating 
#@Baoning Wu, Brian D. Davison
#t2009
#c382
#index745
#id760
#%
#!

#*topical trustrank using web spam web spam behavior attempts search engine ranking algorithms trustrank recent algorithm web spam trustrank sense set used trustrank sufficiently representative cover different topics web given set trustrank larger communities propose use topical information partition set scores topic separately address issues combination scores page used determine ranking experimental results large datasets topical trustrank better performance trustrank spam sites pages compared trustrank best technique decrease spam ranked sites 
#@Baoning Wu, Vinay Goel, Brian D. Davison
#t2009
#c38
#index746
#id761
#%
#!

#*utility scoring product reviews identify new task research text sentiment analysis predicting utility product reviews orthogonal polarity classification opinion extraction build regression models incorporating diverse set features achieve highly competitive performance utility scoring real world data sets 
#@Zhu Zhang, Balaji Varadarajan
#t2009
#c32
#index747
#id762
#%748,393,397,601,142,189,144,145,146,147,749,750,83,751,752,409,725,753,150
#!

#*identifying sources opinions conditional random fields extraction patterns recent systems developed sentiment classification opinion recognition opinion analysis detecting polarity strength aspect opinion analysis identifying sources opinions sentiments view problem information extraction task adopt hybrid approach combines conditional random fields et al variation model source identification sequence tagging task learns extraction patterns results combination methods performs better resulting identifies opinion sources precision recall using noun matching measure precision recall using overlap measure 
#@Yejin Choi, Claire Cardie, Ellen Riloff, Siddharth Patwardhan
#t2009
#c108
#index748
#id763
#%
#!

#*exploiting subjectivity classification improve information extraction information extraction systems false hits variety reasons observed false hits occur sentences contain subjective language opinions sentiments observations explore idea using subjectivity analysis improve precision information extraction systems paper uses subjective sentence classifier filter extractions different strategies using subjectivity classifications including aggressive strategy extractions subjective sentences complex strategies extractions evaluated performance different approaches data set filtering extractions subjective sentences aggressive filtering strategies improved precision minimal recall loss introduction 
#@Ellen Riloff, Janyce Wiebe, William Phillips
#t2009
#c103
#index749
#id764
#%
#!

#*learning subjective nouns using extraction pattern bootstrapping explore idea creating subjectivity classifier uses lists subjective nouns learned bootstrapping algorithms goal research develop subjective sentences objective sentences use bootstrapping algorithms exploit extraction patterns learn sets subjective nouns naive bayes classifier using subjective nouns features subjectivity clues identified prior 
#@Ellen Riloff, Janyce Wiebe, Theresa Wilson
#t2009
#c383
#index750
#id765
#%
#!

#*multi perspective question answering using corpus investigate techniques support answering opinion based questions present corpus opinion questions answers using corpus compare contrast properties fact opinion questions answers based characteristics opinion fact answers argue traditional fact based approaches difficulty setting initial step development systems investigate use machine learning rule based subjectivity opinion source filters used guide systems 
#@Veselin Stoyanov, Claire Cardie, Janyce Wiebe
#t2009
#c108
#index751
#id766
#%
#!

#*bootstrapping method learning semantic using extraction pattern contexts paper describes bootstrapping algorithm called learns high quality semantic multiple categories corpus words semantic category learn new words category semantic class word based information large extraction pattern contexts evaluate semantic categories semantic produced higher precision produced previous techniques categories showing substantial improvement 
#@Michael Thelen, Ellen Riloff
#t2009
#c384
#index752
#id767
#%
#!

#*recognizing contextual polarity phrase level sentiment analysis paper presents new approach phrase level sentiment analysis determines expression polar polarity polar expressions approach able automatically identify contextual polarity large subset sentiment expressions achieving results significantly better baseline 
#@Theresa Wilson, Janyce Wiebe, Paul Hoffmann
#t2009
#c108
#index753
#id768
#%
#!

#*review spam detection common practice commerce web sites enable customers reviews products reviews provide valuable sources information products used potential customers opinions existing users purchase product used product identify problems products competitive intelligence information unfortunately importance reviews gives good spam contains false positive negative opinions paper make attempt study review spam spam detection best knowledge reported study problem 
#@Nitin Jindal, Bing Liu
#t2009
#c38
#index754
#id769
#%644,742,385,755,743,746
#!

#*fast statistical spam filter approximate classifications statistical based bayesian filters popular important spam despite effectiveness processing overhead scaling enterprise level mail servers example dictionary characteristic approach limited memory access rate relatively increases cpu speed address scaling issue technique bayesian filters based approximate classification approximation uses methods hash based encoding approximation based popular filter data structure extension support value retrieval encoding used data structure methods introduce additional errors bayesian approach errors biased false negative classification demonstrate known spam filters achieving identical false positive rate similar false negative rate original filters 
#@Kang Li, Zhenyu Zhong
#t2009
#c385
#index755
#id770
#%
#!

#*sentiment analysis capturing using natural language processing paper sentiment analysis approach extract sentiments associated positive negative specific subjects document instead classifying document positive negative essential issues sentiment analysis identify sentiments expressed texts expressions indicate positive negative opinions subject order improve accuracy sentiment analysis important identify semantic relationships sentiment expressions subject applying semantic analysis syntactic parser sentiment lexicon prototype achieved high precision depending data finding sentiments web pages news articles 
#@Nasukawa Tetsuya, Yi Jeonghee
#t2009
#c386
#index756
#id771
#%145,757,397,358,758,707
#!

#*message understanding conference history recently series message understanding designed evaluate research information extraction introduced prior range different tasks evaluations conducted new discuss results evaluations 
#@Ralph Grishman, Beth Sundheim
#t2009
#c387
#index757
#id772
#%
#!

#*architecture data model present architecture data model robust scalable document analysis framework architecture allowing rapid application development existing language analysis new discuss design issues arise requirements strength efficiency scalability constrained interactions common data model annotation store document vocabulary lexical cache focusing meta annotation based finite state linguistic filters implemented framework component extensively research environments broad range text analysis mining tasks 
#@Mary S. Neff, Roy J. Byrd, Branimir K. Boguraev
#t2009
#c161
#index758
#id773
#%
#!

#*sentiment given topic present sentiment sa sentiment opinion subject online text documents instead classifying sentiment entire document subject sa references given subject determines sentiment references using natural language processing nlp techniques sentiment analysis consists topic specific feature term extraction sentiment extraction subject sentiment association relationship analysis sa utilizes linguistic resources analysis sentiment lexicon sentiment pattern database performance algorithms online product review articles digital music reviews general documents including general news articles 
#@Jeonghee Yi, Tetsuya Nasukawa, Razvan Bunescu, Wayne Niblack
#t2009
#c388
#index759
#id774
#%760,393,761,398,762,763,358,764,765,145,766,767,703,704,409,768,769
#!

#*finding parts large corpora present method extracting parts objects given large corpus method finds words accuracy words ranked list end user added existing wordnet used semantic lexicon 
#@Matthew Berland, Eugene Charniak
#t2009
#c104
#index760
#id775
#%
#!

#*accurate methods statistics work statistical analysis text cases reported literature statistical methods used statistical significance results addressed particular assumptions used leading results assumption normal distribution ability analyze rare events unfortunately rare events make large fraction real text applicable methods based likelihood ratio tests available yield good results relatively small samples tests implemented efficiently used detection terms determination domain specific terms cases measures perform better methods previously used cases traditional table methods work likelihood ratio tests described nearly identical paper describes basis measure based likelihood applied analysis text 
#@Ted Dunning
#t2009
#c279
#index761
#id776
#%
#!

#*sentence processing information access world wide web paper describes start information server built artificial intelligence available world wide web start server provides users access multi media information response questions formulated english years start server millions questions users world start server built foundations sentence level natural language processing provided start natural language idea natural language multi media information segments paper overview sentence level processing start information segments collections english sentences makes possible use power sentence level natural language processing service multi media information access paper world wide web 
#@Boris Katz
#t2009
#c389
#index762
#id777
#%
#!

#*mining open answers data important tools marketing managing customer relationships answers open questions particular contain valuable information provide important basis business summaries human make open answers tend rely 
#@Hang Li, Kenji Yamanishi
#t2009
#c3
#index763
#id778
#%
#!

#*building large annotated corpus english paper review experience constructing large annotated corpus corpus 
#@Mitchell P. Marcus, Mary Ann Marcinkiewicz, Beatrice Santorini
#t2009
#c279
#index764
#id779
#%
#!

#*nouns wordnet lexical definition noun consists immediate followed relative describes instance instances information provides basis organizing files wordnet wordnet lexical connected lexical database hierarchy equal speed 
#@George A. Miller
#t2009
#c235
#index765
#id780
#%
#!

#*maximum entropy model tagging paper presents statistical model corpus annotated assigns previously unseen text state art accuracy model classified maximum entropy model simultaneously uses contextual features predict furthermore paper demonstrates use specialized features model difficult tagging discusses corpus consistency problems discovered implementation 
#@Adwait Ratnaparkhi
#t2009
#c390
#index766
#id781
#%
#!

#*style second advertisements targeted program objective textual analysis incorporated measures style word word used score portion advertisements analysis advertisements directed active longer negative directed comparison advertisements directed age showed linguistic complexity words fewer common words text directed measures used advertisements produced significant difference associated age target population interaction 
#@L. Rovinelli, C. Whissell
#t2009
#c391
#index767
#id782
#%
#!

#*model based feedback modeling approach information retrieval language modeling approach retrieval shown perform empirically advantage new approach statistical foundations feedback important component retrieval new retrieval approach original query usually adding additional terms expansion based feedback interpretation original query paper present principled approach feedback language modeling approach specifically treat feedback updating query language model based evidence carried feedback documents model based feedback strategy easily extension language modeling approach propose evaluate different approaches updating query language model based feedback documents based generative probabilistic model feedback documents based feedback documents experiment results approaches effective outperform rocchio feedback approach 
#@ChengXiang Zhai, J Lafferty
#t2009
#c392
#index768
#id783
#%
#!

#*exact maximum likelihood estimation word mixtures mixture model generating document generative language model used information retrieval using model situations need maximum likelihood estimation density multinomial given fixed mixture weight multinomial 
#@Yi Zhang, Wei Xu, Jamie Callan
#t2009
#c22
#index769
#id784
#%
#!

#*topic sentiment mixture paper define problem topic sentiment analysis weblogs propose novel probabilistic model capture mixture topics sentiments simultaneously proposed topic sentiment mixture tsm model reveal latent topical facets weblog collection results ad hoc query associated sentiments provide general sentiment models applicable ad hoc topics specifically designed hmm structure sentiment models topic models estimated tsm extract topic life sentiment dynamics empirical experiments different weblog datasets approach effective modeling topic facets sentiments extracting dynamics weblog collections tsm model quite general applied text collections mixture topics sentiments potential applications search result summarization opinion tracking user behavior prediction 
#@Qiaozhu Mei, Xu Ling, Matthew Wondra, Hang Su, ChengXiang Zhai
#t2009
#c393
#index770
#id785
#%176,748,179,180,771,199,772,184,773,142,187,774,188,775,776,189,144,145,318,777,778,779,595
#!

#*clustering versus categories information exploration information express user interface search results meaningful groups order help make sense results help decide study provided ability group search results search response having grouping mechanism available open research questions generate useful design interfaces support exploration using grouping currently methods quite popular clustering categorization approaches summarize advantages based results studies 
#@Marti A. Hearst
#t2009
#c7
#index771
#id786
#%
#!

#*inference process area linguistics concerned internal structure words information retrieval generally attention word structure account variability word forms use paper experiments determine importance effect performance role analysis word sense disambiguation identifying lexical 
#@Robert Krovetz
#t2009
#c394
#index772
#id787
#%
#!

#*allocation structured mixture models topic correlations latent dirichlet allocation lda related topic models increasingly popular tools summarization discovery discrete data lda does capture correlations topics paper introduce allocation model captures arbitrary nested possibly sparse correlations topics using directed graph represent individual words vocabulary interior node represents correlation words interior nodes topics provides alternative recent work captures correlations pairs topics using text data research paper corpora improved performance document classification likelihood held data ability support topics topical keyword 
#@Wei Li, A McCallum
#t2009
#c22
#index773
#id788
#%
#!

#*discovering evolutionary theme patterns text exploration temporal text mining temporal text mining concerned discovering temporal patterns text information collected time text information time applications multiple domains summarizing events news articles research trends scientific literature paper study particular task discovering summarizing evolutionary patterns themes text stream define new text mining problem present general probabilistic methods solving problem discovering latent themes text constructing evolution graph themes analyzing life themes evaluation proposed methods different domains news articles literature shows proposed methods discover interesting evolutionary theme patterns effectively 
#@Qiaozhu Mei, ChengXiang Zhai
#t2009
#c3
#index774
#id789
#%
#!

#*tools blog analysis demonstrate tracking analyzing largest community thousands blog posts performs various analyses posts presents results 
#@G Mishne, M de Rijke
#t2009
#c103
#index775
#id790
#%
#!

#*predicting movie sales sentiment volume discussion product weblogs recently shown products performance paper study applying sentiment analysis methods weblog data results better correlation volume domain movies main finding positive sentiment better predictor movie success applied limited context references movie weblogs prior release 
#@G Mishne, N Glance
#t2009
#c103
#index776
#id791
#%
#!

#*estimation mixture models robust pseudo relevance feedback pseudo relevance feedback proven effective strategy improving retrieval accuracy retrieval models performance existing pseudo feedback methods affected significantly parameters number feedback documents use relative weight original query terms parameters generally set error paper present robust method pseudo feedback based statistical language models main idea integrate original query feedback documents single probabilistic mixture model estimation language model parameters model information feedback documents added original query unlike existing feedback methods new method parameter experiment results representative data sets new method significantly robust state art baseline language modeling approach feedback comparable better retrieval accuracy 
#@Tao Tao, ChengXiang Zhai
#t2009
#c35
#index777
#id792
#%
#!

#*expressions opinions language paper describes corpus annotation project study issues manual annotation opinions sentiments evaluations private states language resulting corpus annotation scheme described examples use addition manual annotation process results inter agreement study sentence corpus articles world presented 
#@Janyce Wiebe, Theresa Wilson, Claire Cardie
#t2009
#c395
#index778
#id793
#%
#!

#*sentiment extracting sentiments given topic using natural language processing techniques present sentiment sa sentiment opinion subject online text documents instead classifying sentiment entire document subject sa references given subject determines sentiment references using natural language processing nlp techniques sentiment analysis consists topic specific feature term extraction sentiment extraction subject sentiment association relationship analysis sa utilizes linguistic resources analysis sentiment lexicon sentiment pattern database performance algorithms online product review articles digital music reviews general documents including general news articles 
#@Jeonghee Yi, Tetsuya Nasukawa, Razvan C. Bunescu, Wayne Niblack
#t2009
#c14
#index779
#id794
#%
#!

#*using polarity scores words sentence level opinion extraction opinion analysis task study task contains challenges opinion sentence extraction opinion polarity opinion holder extraction relevance sentence extraction new tasks proven trec paper introduce analyzing opinionated information proposed decide opinion words process opinion sentences operators taken consideration opinion polarity opinion operators used clues locations opinion performance opinion extraction polarity achieves measure metric metric second best 
#@Lun-Wei Ku, Yong-Sheng Lo, Hsin-Hsi Chen
#t2009
#c396
#index780
#id795
#%390,393,385,139,721,145,147,781,782,783,784,750,141
#!

#*extracting semantic orientations words using model propose method extracting semantic orientations words desirable semantic orientations use mean field approximation compute approximate probability function instead intractable actual probability function propose criterion parameter selection basis given small number words proposed method semantic orientations high accuracy experiments english lexicon result comparable best value reported 
#@Hiroya Takamura, Takashi Inui, Manabu Okumura
#t2009
#c104
#index781
#id796
#%
#!

#*identify recognizing opinions subjectivity natural language refers aspects language used express opinions evaluations 
#@Janyce Wiebe, Theresa Wilson, Matthew Bell
#t2009
#c369
#index782
#id797
#%
#!

#*overview trec track track introduced trec given trec topic ordered list documents systems relevant novel sentences user set task integrates aspects retrieval information filtering year using trec topics documents developed new topics specifically track topics classes events opinions documents 
#@Ian Soboroff, Donna Harman
#t2009
#c52
#index783
#id798
#%
#!

#*learning models extracting consumer opinions extracting sentiments unstructured text important problem accurate method enable example online opinions internet learn customers preferences economic marketing research advantage paper propose stage bayesian algorithm able capture dependencies words time finds vocabulary efficient purpose extracting sentiments experimental results movie reviews data set algorithm able select feature set substantially fewer predictor variables data set leads better predictions sentiment orientations state art machine learning methods suggest sentiments conditional relations words keywords high frequency words 
#@Xue Bai, Rema Padman, Edoardo Airoldi
#t2009
#c142
#index784
#id799
#%
#!

#*clustering methods collaborative filtering grouping people clusters based items allows accurate recommendations new items purchase movies movies like items based similarity collaborative filtering domains books movies does work data sparse given person seen small fraction movies accurate predictions grouping people clusters similar movies grouping movies clusters tend people finding optimal clusters movie groups used help determine people groups present formal statistical model collaborative filtering compare different algorithms estimating model parameters including variations means clustering sampling formal model easily extended handle clustering objects multiple attributes 
#@Lyle H. Ungar, Dean Foster, Ellen Andre, Star Wars, Fred Star Wars, Dean Star Wars, Jason Hiver Whispers
#t2009
#c397
#index785
#id800
#%786,787
#!

#*using collaborative filtering information experimental mail developed research center belief information filtering effective humans involved filtering process designed support content based filtering collaborative filtering people help perform filtering documents read called filters intended handle stream electronic documents serves mail filter components document store annotation store little reader browser server architecture various components query language described 
#@David Goldberg, David A. Nichols, Brian M. Oki, Douglas B. Terry
#t2009
#c7
#index786
#id801
#%
#!

#*applying collaborative filtering news project designed implemented evaluated collaborative filtering high volume high discussion list service internet individual discussion hundreds messages theory organization allows readers select content interests practice 
#@Joseph A. Konstan, Bradley N. Miller, David Maltz, Jonathan L. Herlocker, Lee R. Gordon, John Riedl
#t2009
#c7
#index787
#id802
#%
#!

#*collaborative filtering gaussian probabilistic latent semantic analysis collaborative filtering aims learning predictive models user preferences interests behavior community data database available user preferences paper new model based algorithm designed task based generalization probabilistic latent semantic analysis continuous response variables specifically assume observed user ratings modeled mixture user communities groups users groups community characterized gaussian distribution normalized ratings item ratings performed user specific manner account variations variance ratings experiments eachmovie data set proposed approach compares collaborative filtering techniques 
#@Thomas Hofmann
#t2009
#c35
#index788
#id803
#%786,789,787
#!

#*open architecture collaborative filtering collaborative filters help people make choices based opinions people collaborative filtering help people articles like stream available articles news reader predicted scores make easy users rate articles read rating servers called better gather ratings rating servers predict scores based heuristic people 
#@Paul Resnick, Neophytos Iacovou, Mitesh Suchak, Peter Bergstrom, John Riedl
#t2009
#c398
#index789
#id804
#%
#!

#*collaborative filtering models preferences ratings paper new model collaborative filtering work fact users similar preferences items different rating schemes example user tend assign higher rating items user unlike previous models collaborative filtering determine similarity users based rating performance model users preferences items separately users rating scheme specifically user build separate models preference model capturing items user rating model capturing user rate item given preference information similarity users computed based underlying preference model instead surface ratings compare new model representative previous approaches data sets experiment results new model outperforms previous approaches tested consistently data sets 
#@Rong Jin, Luo Si, ChengXiang Zhai, James P. Callan
#t2009
#c32
#index790
#id805
#%791,789,792,793,794,795,796
#!

#*collaborative filtering generalized vector space model collaborative filtering technique documents users based similar users users tend like recommend documents generalized vector space model information retrieval represents document vector similarities documents process collaborative filtering nearly identical process retrieval using matrix user ratings using observation model filtering using document content possible 
#@Ian Soboroff, Charles K. Nicholas
#t2009
#c35
#index791
#id807
#%
#!

#*combining content collaboration text filtering technique combining input document content text technique uses latent indexing create collaborative view collection user pro pro term vectors constructed relevant user tion need using standard text collections approach performs quite compared approaches 
#@Ian Soboroff, Charles K. Nicholas
#t2009
#c399
#index792
#id809
#%
#!

#*content collaborative filtering improved recommendations recommender systems use collaborative filtering content based methods predict new items user methods advantages provide good recommendations situations incorporating components methods hybrid recommender paper present effective framework combining content collaboration approach uses content based predictor enhance existing user data provides personalized collaborative filtering present experimental results approach content collaborative filtering performs better pure content based predictor pure collaborative filter naive hybrid approach 
#@Prem Melville, Raymond J. Mooney, Ramadass Nagarajan
#t2009
#c175
#index793
#id810
#%
#!

#*framework collaborative filtering algorithm development evaluation present based framework shared building studying collaborative filtering systems consists components prediction engine evaluation visualization component prediction engine provides common interface implementing different prediction algorithms evaluation provides testing methodology metrics analyzing accuracy run time performance prediction algorithms visualization component suggests graphical representations development analysis prediction algorithms demonstrate eachmovie data set comparing prediction algorithms traditional correlation based method support vector machines new accurate scalable correlation based method based clustering techniques 
#@Danyel Fisher, Kris Hildrum, Jason I. Hong, Mark W. Newman, Megan Thomas, Rich Vuduc
#t2009
#c35
#index794
#id811
#%
#!

#*maximum likelihood data em algorithm applicable algorithm computing maximum likelihood estimates data presented various levels theory showing behaviour likelihood convergence algorithm derived examples including missing value situations applications data finite mixture models variance component estimation estimation factor analysis 
#@A. P. Dempster, N. M. Laird, D. B. Rubin
#t2009
#c400
#index795
#id812
#%
#!

#*collaborative filtering personality diagnosis hybrid memory model based approach growth internet commerce use collaborative filtering cf algorithms recommender systems systems knowledge known preferences multiple users recommend items users cf methods make recommendations items web pages movies books researchers proposed evaluated approaches generating recommendations evaluate new method called personality diagnosis pd given users preferences items compute probability personality type users turn probability like new items pd advantages traditional similarity weighting techniques data prediction new data added easily pd meaningful probabilistic interpretation explain augment results report empirical results eachmovie database movie ratings user profile data collected digital library science research papers probabilistic framework naturally supports variety particular consider value information computation 
#@DM Pennock, Eric Horvitz, S Lawrence, C. Lee Giles
#t2009
#c129
#index796
#id813
#%
#!

#*effective missing data prediction collaborative filtering memory based collaborative filtering algorithms widely adopted popular recommender systems approaches data sparsity poor prediction quality problems usually user item matrix quite sparse directly leads recommendations paper focuses memory based collaborative filtering problems crucial factors similarity computation users items missing data prediction algorithms use enhanced correlation algorithm adding parameter potential decrease accuracy computing similarity users items second propose effective missing data prediction algorithm information users items taken account algorithm set similarity threshold users items respectively prediction algorithm determine predicting missing data address predict missing data combination user item information finally empirical studies dataset shown proposed method outperforms art collaborative filtering algorithms robust data sparsity 
#@Hao Ma, Irwin King, Michael R. Lyu
#t2009
#c35
#index797
#id814
#%798,799,800,801,802,788,803,804,805,806,807,796,808,789,809,810,785,811,812
#!

#*unifying collaborative content based filtering collaborative content based filtering applied context recommender systems user preference prediction paper proposes novel unified approach systematically integrates available training information past user item ratings attributes items users learn prediction function key method design suitable similarity function user item pairs allows generalization user item dimensions propose line algorithm learning experimental results eachmovie data set significant improvements standard approaches 
#@Justin Basilico, Thomas Hofmann
#t2009
#c22
#index798
#id815
#%
#!

#*collaborative filtering privacy factor analysis collaborative filtering cf valuable commerce direct recommendations music movies news today systems including privacy computing great potential individuals share kinds information buy privacy paper new method collaborative filtering privacy individual data method based probabilistic factor analysis model privacy protection provided peer peer protocol described paper factor analysis approach missing data requiring values experiments suggest accurate method cf date new algorithm advantages speed storage previous algorithms finally suggest applications approach kinds statistical analyses survey data 
#@John F. Canny
#t2009
#c35
#index799
#id816
#%
#!

#*combining content based collaborative filters online explosive growth lists web sites news effective filtering solutions collaborative filtering combines informed opinions humans make personalized accurate predictions content based filtering uses speed computers make complete fast predictions work present new filtering approach combines coverage speed content filters depth collaborative filtering apply research approach online filters useful wide news present design filtering results preliminary experiments suggest approach 
#@Mark Claypool, Anuja Gokhale, Tim Miranda, Pavel Murnikov, Dmitry Netes, Matthew Sartin
#t2009
#c401
#index800
#id817
#%
#!

#*item based recommendation explosive growth world wide web commerce led development recommender systems personalized information filtering technology used identify set items certain user user based collaborative filtering successful technology building recommender systems date extensively used commercial recommender systems unfortunately computational complexity methods grows number customers typical commercial applications millions address scalability concerns model based recommendation techniques developed techniques analyze user item matrix discover relations different items use relations compute list recommendations article present class model based recommendation algorithms determines similarities various items uses identify set items recommended key steps class algorithms method used compute similarity items method used combine similarities order compute similarity items candidate recommender item experimental evaluation real datasets shows item based algorithms orders magnitude faster traditional user neighborhood based recommender systems provide recommendations comparable better quality 
#@M Deshpande, G Karypis
#t2009
#c402
#index801
#id818
#%
#!

#*empirical analysis design choices neighborhood based collaborative filtering algorithms collaborative filtering systems predict user new items based recommendations people similar interests instead performing content indexing content analysis collaborative filtering systems rely ratings community predictions based human ratings collaborative filtering systems potential provide filtering based complex attributes quality implementations collaborative filtering apply variation neighborhood based prediction algorithm variations similarity metrics weighting approaches combination measures rating implementation parameters choice technique appropriate situations significant effect accuracy parameter person implementing collaborative filtering make hard design choices little article provides set recommendations guide design neighborhood based prediction systems based results empirical study apply analysis framework neighborhood based prediction approach components examines variants key parameters component components identified similarity computation neighbor selection rating combination 
#@Jonathan L. Herlocker, Joseph A. Konstan, John Riedl
#t2009
#c206
#index802
#id819
#%
#!

#*latent semantic models collaborative filtering collaborative filtering aims learning predictive models user preferences interests behavior community data database available user preferences article new family model based algorithms designed task algorithms rely statistical modelling technique introduces latent class variables mixture model setting discover user communities investigate variations deal discrete continuous response variables different objective functions main advantages technique standard memory based methods higher accuracy constant time prediction explicit compact model representation used user experimental evaluation shows substantial improvements existing methods published results obtained 
#@Thomas Hofmann
#t2009
#c37
#index803
#id820
#%
#!

#*automatic weighting scheme collaborative filtering collaborative filtering identifies information particular user based information provided similar users memory based approaches collaborative filtering correlation approach identify similarity users comparing ratings set items approaches different items weighted equally functions impact rating different users taken consideration example item highly users smaller impact user similarity item different types users tend different ratings simple weighting methods variance weighting address problem empirical studies shown improving performance collaborative filtering paper present optimization algorithm automatically compute weights different items based ratings training users specifically new weighting scheme create clustered distribution user vectors item space users similar interests separating users different interests empirical studies datasets shown new weighting scheme substantially improves performance correlation method collaborative filtering 
#@Rong Jin, Joyce Y. Chai, Luo Si
#t2009
#c35
#index804
#id821
#%
#!

#*clustering collaborative filtering applications collaborative filtering systems users identify items providing predictions based ratings users quality predictions depends strongly available ratings collaborative filtering algorithms perform ratings available paper identify important situations sparse ratings bootstrapping collaborative filtering users providing recommendations new users 
#@Arnd Kohrs, Bernard Merialdo
#t2009
#c403
#index805
#id822
#%
#!

#*amazon com recommendations item item collaborative filtering recommendation algorithms best known use commerce web sites use input customers interests generate list recommended items applications use items customers purchase explicitly rate represent interests use attributes including items viewed data subject interests amazon com use recommendation algorithms online store customer store changes based customer interests showing programming software new rates important measures web based email effectiveness content advertisements lists 
#@Greg Linden, Brent Smith, Jeremy York
#t2009
#c338
#index806
#id823
#%
#!

#*collaborative filtering algorithm evaluation metric accurately model user experience collaborative filtering cf systems tool deal information heart systems algorithms generate predictions recommendations article empirically demonstrate cf recommendation algorithms result dramatically user experience response introduce new belief distribution algorithm provides substantially user modeling belief distribution algorithm nearest neighbor algorithms performed past produces predictions belief distributions rating values point rating value addition illustrate use mean error metric long propose use modified precision metric accurately evaluating user experience 
#@Matthew R. McLaughlin, Jonathan L. Herlocker
#t2009
#c35
#index807
#id824
#%
#!

#*fast maximum matrix factorization collaborative prediction maximum matrix factorization recently suggested et al dimensional alternative low rank approximations standard factor models formulated semi programming learned using standard current handle problems matrices dimensionality investigate direct gradient based optimization method demonstrate large collaborative prediction problems compare results obtained substantially outperforms methods tested 
#@Jason D. M. Rennie, Nathan Srebro
#t2009
#c22
#index808
#id825
#%
#!

#*item based collaborative filtering recommendation algorithms recommender systems apply knowledge discovery techniques problem making personalized recommendations information products services interaction systems especially nearest neighbor collaborative filtering based ones achieving widespread success web tremendous growth available information number web sites recent years key challenges recommender systems producing high quality recommendations performing recommendations second millions users items achieving high coverage face data sparsity traditional collaborative filtering systems work increases number new recommender technologies needed quickly produce high quality recommendations large scale problems address issues explored item based collaborative filtering techniques item based techniques analyze user item matrix identify relationships different items use relationships compute recommendations users paper analyze different item based recommendation generation algorithms look different techniques computing item item similarities item item correlation similarities item vectors different techniques obtaining recommendations weighted sum regression model finally experimentally evaluate results compare basic nearest neighbor approach experiments suggest item based algorithms provide dramatically better performance user based algorithms time providing better quality best available algorithms 
#@Badrul M. Sarwar, G Karypis, Joseph A. Konstan, John Riedl
#t2009
#c38
#index809
#id826
#%813,814,815,786,816,817,787,789,818,819,820,785
#!

#*flexible mixture model collaborative filtering paper presents flexible mixture model collaborative filtering existing partitioning clustering algorithms collaborative filtering clustering users items simultaneously assuming user item single cluster furthermore introduction preference nodes proposed framework able explicitly model users rate items vary dramatically users similar items empirical study datasets movie ratings shown new algorithm outperforms collaborative filtering algorithms substantially 
#@Luo Si, Rong Jin
#t2009
#c22
#index810
#id827
#%
#!

#*unifying user based item based collaborative filtering approaches similarity memory based methods collaborative filtering predict new ratings weighted ratings respectively pairs similar users items practice large number ratings similar users similar items available sparsity rating data prediction quality poor paper memory based collaborative filtering problem generative probabilistic framework individual user item ratings predictors missing ratings final rating estimated predictions sources predictions based ratings item users predictions based different item ratings user ratings predicted based data similar users rating similar items existing user based item based approaches correspond simple cases framework complete model robust data sparsity different types ratings used additional ratings similar users similar items employed background model predictions experiments demonstrate proposed methods robust data sparsity better recommendations 
#@Jun Wang, Arjen P. de Vries, Marcel J. T. Reinders
#t2009
#c35
#index811
#id828
#%
#!

#*scalable collaborative filtering using cluster based smoothing memory based approaches collaborative filtering identify similarity users comparing ratings set items past memory based approaches shown fundamental problems data sparsity difficulty scalability model based approaches proposed problems approaches tends limit range users paper present novel approach combines advantages kinds approaches smoothing based method approach clusters generated training data provide basis data smoothing neighborhood selection result provide higher accuracy increased efficiency recommendations empirical studies datasets eachmovie new proposed approach consistently outperforms state art collaborative filtering algorithms 
#@Gui-Rong Xue, Chenxi Lin, Qiang Yang, Wensi Xi, Hua-Jun Zeng, Yong Yu, Zheng Chen
#t2009
#c35
#index812
#id829
#%821,822,800,794,823,824,825,805,796,789,791,785,826
#!

#*new graph theoretic approach collaborative filtering paper introduces new novel approach collaborative filtering new technique appropriate commerce groups relatively homogeneous items compact videos books software like contrast known collaborative filtering techniques new algorithm graph theoretic based new concepts demonstrated paper technique fast scalable accurate requires learning makes use hierarchical classification scheme order introduce context rating process uses called links order items recommend items cross group new technique key engines intelligent recommendation algorithm project developed ibm research addition recommendation engines contains situation determine appropriate engines particular commerce engine optimizing advertisements 
#@Charu C. Aggarwal, Joel L. Wolf, Kun-Lung Wu, Philip S. Yu
#t2009
#c3
#index813
#id830
#%
#!

#*recommendation classification using social content based information recommendation recommendation systems make artifacts user instance predict user interested particular movie social methods collect ratings artifacts individuals use nearest neighbor techniques make recommendations user concerning new artifacts methods use significant information available nature lists 
#@Chumki Basu, Haym Hirsh, William W. Cohen
#t2009
#c175
#index814
#id831
#%
#!

#*learning collaborative information filters predicting items user like basis users ratings items established strategy adopted recommendation services internet seen classification problem algorithms proposed far draw results machine learning literature propose representation collaborative filtering tasks allows application machine learning algorithm identify current collaborative filtering techniques propose use learning algorithms feature extraction techniques specifically address limitations previous approaches best performing algorithm based singular value decomposition initial matrix user ratings exploiting latent structure essentially need users rate common items order predictors preferences evaluate proposed algorithm large database user ratings approach significantly outperforms current collaborative filtering algorithms 
#@D Billsus, M Pazzani
#t2009
#c22
#index815
#id832
#%
#!

#*combining collaborative filtering personal agents better recommendations information filtering agents collaborative filtering attempt information identifying items user information filtering focuses analysis item content development personal user profile collaborative filtering cf focuses identification users similar use opinions recommend items technique advantages limitations suggest combined paper shows cf framework used combine personal agents opinions community users produce better recommendations agents users produce shows using cf create personal combination set agents produces better results individual agents combination mechanisms key results users avoid having select agents use let cf framework select best ones 
#@Nathaniel Good, J. Ben Schafer, Joseph A. Konstan, Al Borchers, Badrul M. Sarwar, Jonathan L. Herlocker, John Riedl
#t2009
#c175
#index816
#id833
#%
#!

#*evaluation item based recommendation algorithms explosive growth world wide web commerce led development recommender systems personalized information filtering technology used identify set items certain user user based collaborative filtering successful technology building recommender systems date extensively used commercial recommender systems unfortunately computational complexity methods grows 
#@G Karypis
#t2009
#c32
#index817
#id834
#%
#!

#*recommender systems recommender systems augment natural social process typical recommender people provide recommendations inputs appropriate cases primary 
#@Hannes Werthner, Hans Robert Hansen, Francesco Ricci
#t2009
#c142
#index818
#id835
#%
#!

#*using filtering agents improve prediction quality research collaborative filtering collaborative filtering systems help address information using opinions users community make personal recommendations documents user collaborative filtering systems user opinions relative large number documents available sparsity problem reduce utility filtering reducing number documents make recommendations quality recommendations paper defines model integrating content based ratings collaborative filtering model allows collaborative filtering systems address sparsity strength content filtering techniques identify evaluate metrics assessing effectiveness specifically filtering enhancements general finally experimentally approach showing simple increase utility users collaborative filtering systems 
#@Badrul M. Sarwar, Joseph A. Konstan, Al Borchers, Jonathan L. Herlocker, Bradley N. Miller, John Riedl
#t2009
#c398
#index819
#id836
#%
#!

#*recommender systems commerce recommender systems changing used commerce sites business tools world commerce largest commerce web sites using recommender systems help customers products purchase recommender learns customer products valuable available products paper present recommender systems help commerce sites increase sales analyze sites use recommender systems including sites use recommender based examples create recommender systems including interfaces present customers technologies used create recommendations inputs need customers conclude ideas new applications recommender systems commerce 
#@J. Ben Schafer, Joseph A. Konstan, John Riedl
#t2009
#c404
#index820
#id837
#%
#!

#*content based collaborative recommendation online readers need tools help content available world wide web traditional media readers provided making includes implicit form explicit form recommendation services movie reviews electronic medium offers new create recommendation services ones time track interests recommendation web versions 
#@Marko Balabanovic, Yoav Shoham
#t2009
#c7
#index821
#id839
#%
#!

#*class based gram models natural language address problem predicting word previous words sample text particular discuss gram models based classes words discuss statistical algorithms assigning words classes based frequency occurrence words able extract classes based semantically based depending nature underlying statistics 
#@Peter F. Brown, Vincent J. Della Pietra, Peter V. de Souza, Jennifer C. Lai, Robert L. Mercer
#t2009
#c279
#index822
#id840
#%
#!

#*eigentaste constant time collaborative filtering algorithm eigentaste collaborative filtering algorithm uses queries real user ratings common set items applies principal component analysis pca resulting dense subset ratings matrix pca dimensionality reduction clustering users rapid computation recommendations database users standard nearest neighbor techniques require processing time compute recommendations eigentaste requires constant time compare eigentaste alternative algorithms using data online collected approximately ratings users use normalized mean error measure compare performance different algorithms use uniform normal distribution models derive estimates predictions random dataset eigentaste recommendations orders magnitude faster loss accuracy online eigentaste 
#@Kenneth Y. Goldberg, Theresa Roeder, Dhruv Gupta, Chris Perkins
#t2009
#c206
#index823
#id841
#%
#!

#*evaluating collaborative filtering recommender systems recommender systems evaluated ways article review key evaluating collaborative filtering recommender systems user tasks evaluated types analysis datasets used ways prediction quality measured evaluation prediction attributes quality user based evaluation addition evaluation strategies used prior researchers present empirical results analysis various accuracy metrics content domain tested metrics roughly equivalence classes metrics class strongly correlated metrics different classes 
#@Jonathan L. Herlocker, Joseph A. Konstan, Loren G. Terveen, John Riedl
#t2009
#c37
#index824
#id842
#%
#!

#*applying retrieval techniques sparsity problem collaborative filtering recommender systems widely applied application suggest products services information items potential consumers collaborative filtering successful recommendation approach makes recommendations based past transactions feedback consumers sharing similar interests major problem collaborative filtering sparsity problem refers situation feedback data sparse identify similarities consumer interests article propose deal sparsity problem applying retrieval framework related activation algorithms explore associations consumers past transactions feedback associations valuable source information help infer consumer interests explored deal sparsity problem evaluate effectiveness approach conducted experimental study using data set online activation algorithms including constrained algorithm bound search algorithm net parallel search algorithm algorithms compared collaborative filtering approaches consider associations simple graph search approach variations user based approach item based approach experimental results indicate activation based approaches significantly collaborative filtering methods measured recommendation precision recall measure rank score observed activation effect activation approach incorporating associations past data sparse data used infer user preferences lead recommendation performance 
#@Zan Huang, Hsinchun Chen, Daniel Dajun Zeng
#t2009
#c37
#index825
#id843
#%
#!

#*similarity measure instance selection collaborative filtering collaborative filtering successful research applications information filtering commerce nearest neighbor method popular way key technique nearest given user predict interests method fundamental problems sparsity scalability paper present solutions problems adopt techniques matrix method similarity measure instance selection method present improved collaborative filtering algorithm based methods contrast existing collaborative algorithms method shows accuracy performance 
#@Chun Zeng, Chun-Xiao Xing, Li-Zhu Zhou
#t2009
#c38
#index826
#id844
#%877,841,878,879,787,840,821,800,880,881,789,882,883
#!

#*ranking oriented approach collaborative filtering recommender able suggest items likely user systems degree preference represented rating score given database users past ratings set items traditional collaborative filtering algorithms based predicting potential ratings user assign items ranked predicted ratings produce list recommended items paper propose collaborative filtering approach addresses item ranking problem directly modeling user preferences derived ratings measure similarity users based correlation rankings items rating values propose new collaborative filtering algorithms ranking items based preferences similar users experimental results real world movie rating data sets proposed approach outperforms traditional collaborative filtering algorithms significantly measure evaluating ranked results 
#@Nathan N. Liu
#t2009
#c405
#index827
#id845
#%828,829,830,823,831,802,803,832,833,790,834,787,806,797,835,789,809,812
#!

#*learning rank using gradient investigate using gradient methods learning ranking functions propose simple probabilistic cost function introduce implementation ideas using neural network model underlying ranking function present test results data data commercial internet search engine 
#@Christopher J. C. Burges, Tal Shaked, Erin Renshaw, Ari Lazier, Matt Deeds, Nicole Hamilton, Gregory N. Hullender
#t2009
#c22
#index828
#id846
#%
#!

#*learning order applications desirable order classify instances consider problem learning order given feedback form preference judgments effect instance ranked outline stage approach learns conventional means preference function form indicates rank new instances ordered maximize learned preference function problem finding ordering best preference function np complete assumptions simple algorithm guaranteed good approximation discuss line learning algorithm based algorithm finding good linear combination ranking experts use ordering algorithm combined line learning algorithm combination search experts domain specific query expansion strategy www search engine present experimental results demonstrate approach 
#@William W. Cohen, Robert E. Schapire, Yoram Singer
#t2009
#c406
#index829
#id847
#%
#!

#*efficient boosting algorithm combining preferences study problem learning accurately rank set objects combining given collection ranking preference functions problem combining preferences arises applications combining results different search engines problem ranking movies user based movie rankings provided users work begin presenting formal framework general problem analyze efficient algorithm called combining preferences based boosting approach machine learning theoretical results describing algorithms behavior training data new test data seen training efficient implementation algorithm particular restricted common case discuss experiments carried assess performance experiment used algorithm combine different web search strategies query expansion given domain second experiment collaborative filtering task making movie recommendations 
#@Yoav Freund, Raj D. Iyer, Robert E. Schapire, Yoram Singer
#t2009
#c128
#index830
#id848
#%
#!

#*topic sensitive pagerank original pagerank algorithm improving ranking search query results single pagerank vector computed using link structure web capture relative importance pages independent particular search query yield accurate search results propose computing set pagerank vectors biased using set representative topics capture accurately notion importance respect particular topic using biased pagerank vectors generate query specific importance scores pages query time generate accurate rankings single generic pagerank vector keyword search queries compute topic sensitive pagerank scores pages satisfying query using topic query keywords searches context search query performed words web page compute topic sensitive pagerank scores using topic context query 
#@T. Haveliwala
#t2009
#c50
#index831
#id850
#%836,837,838,839,833,170,655
#!

#*gain based evaluation ir techniques modern large retrieval environments tend users large output documents equal relevance users highly relevant documents identified ranked order develop ir techniques direction necessary develop evaluation approaches methods credit ir methods ability retrieve highly relevant documents extending traditional evaluation methods recall precision based binary relevance judgments relevance judgments novel measures based relevance judgments developed article proposes novel measures compute gain user examining retrieval result given ranked position relevance scores retrieved documents ranked result list second similar applies factor relevance scores order retrieved documents relative ideal performance ir techniques based gain able yield novel measures defined discussed use demonstrated case study using trec data sample run results queries trec relevance base used novel relevance judgments point scale test results indicate proposed measures credit ir methods ability retrieve highly relevant documents allow testing statistical significance effectiveness differences graphs based measures provide insight performance ir techniques allow interpretation example user point view 
#@Kalervo Jarvelin, Jaana Kekalainen
#t2009
#c37
#index832
#id851
#%
#!

#*scaling personalized web search recent web search techniques augment traditional text matching global notion importance based linkage structure web google pagerank algorithm searches global notion importance specialized create personalized views importance example importance scores biased according user specified set interesting pages computing storing possible personalized views advance computing personalized views query time computation view requires iterative computation web graph present new graph theoretical results new technique based results personalized views partial vectors partial vectors shared multiple personalized views computation storage costs scale number views approach enables incremental computation construction personalized views partial vectors practical query time present efficient dynamic programming algorithms computing partial vectors algorithm constructing personalized views partial vectors experimental results effectiveness scalability techniques 
#@Glen Jeh, Jennifer Widom
#t2009
#c38
#index833
#id852
#%
#!

#*optimizing search engines using clickthrough data paper presents approach automatically optimizing retrieval quality search engines using clickthrough data intuitively good information retrieval present relevant documents high ranking relevant documents following previous approaches learning retrieval functions examples exist typically require training data generated relevance judgments experts makes difficult expensive apply goal paper develop method utilizes clickthrough data training query log search engine connection log links users presented ranking clickthrough data available low cost taking support vector machine svm approach paper presents method learning retrieval functions theoretical perspective method shown risk framework furthermore shown feasible large sets queries features theoretical results experiment shows method effectively retrieval function meta search engine particular group users google terms retrieval quality training examples 
#@Thorsten Joachims
#t2009
#c3
#index834
#id853
#%830,840,841,842
#!

#*collaborative filtering personality diagnosis hybrid memory model based approach growth internet commerce use collaborative filtering cf algorithms recommender systems systems knowledge known preferences multiple users recommend items users cf methods make recommendations items web pages movies books researchers proposed evaluated approaches generating recommendations evaluate new method called personality diagnosis pd given users preferences items compute probability personality type users turn probability like new items pd advantages traditional similarity weighting techniques data prediction new data added easily pd meaningful probabilistic interpretation explain augment results report empirical results eachmovie database movie ratings user profile data collected digital library science research papers probabilistic framework naturally supports variety particular consider value information computation 
#@DM Pennock, Eric Horvitz, S Lawrence, C. Lee Giles
#t2009
#c129
#index835
#id854
#%
#!

#*improved algorithms topic hyperlinked environment paper addresses problem topic given typical user query quality documents related query topic connectivity analysis shown useful identifying high quality pages topic specific graph hyperlinked documents approach augment previous connectivity analysis based algorithm content analysis identify problems existing approach algorithms results user evaluation reported improvement precision documents pure connectivity analysis 
#@Krishna Bharat, Monika Rauch Henzinger
#t2009
#c35
#index836
#id856
#%843,837,844,844,845,170,846,847,655,848
#!

#*automatic resource analyzing structure associated text design evaluation arc automatically list authoritative web resources sufficiently broad topic goal arc resource lists similar provided yahoo fundamental difference services construct lists manually combination human automated effort arc operates fully automatically evaluation arc yahoo resource lists human users evaluation suggests resources arc frequently better lists resources manually compiled classified topic provide examples arc resource lists reader examine 
#@Soumen Chakrabarti, Byron Dom, Prabhakar Raghavan, S. Rajagopalan, D. Gibson, Jon M. Kleinberg
#t2009
#c46
#index837
#id858
#%
#!

#*search context concept keyword based search engines widespread use today popular means web based information retrieval systems simple considerable required order satisfy non information needs paper presents new paradigm performing search context largely search process providing non users highly relevant results paradigm implemented practice search text query user document views text query document context context driven information retrieval process involves semantic keyword extraction clustering automatically generate new queries submitted host general domain specific search engines search results semantically using context experimental results using context guide search effectively offers users advanced search tool web 
#@Lev Finkelstein, Evgeniy Gabrilovich, Yossi Matias, Ehud Rivlin, Zach Solan, Gadi Wolfman, Eytan Ruppin
#t2009
#c37
#index838
#id860
#%
#!

#*efficient computation pagerank paper discusses efficient techniques computing pagerank ranking metric hypertext documents pagerank computed large web hundreds millions nodes machines limited main memory running time various memory presented pagerank computation million page discuss methods analyzing convergence pagerank based ordering pages present convergence results determining number necessary achieve useful pagerank assignment presence search queries 
#@T. Haveliwala
#t2009
#c83
#index839
#id861
#%
#!

#*tour guide world wide web explore notion tour guide software agent users browsing world wide web web tour guide agent provides similar provided human tour guide user appropriate path collection based knowledge user interests location relevance various items collection way collection past paper describes simple tour guide called given people browsing science web pages users page page suggests appropriate learns experience improve learning algorithms used experimental results showing effectiveness learned case study web tour guide agents 
#@Thorsten Joachims, Dayne Freitag, Tom Mitchell
#t2009
#c407
#index840
#id864
#%
#!

#*agent web browsing user interface agent user browsing world wide web user operates conventional web browser agent tracks user behavior attempts items autonomous exploration links user current position agent browsing strategy consisting best search heuristics user browsing behavior introduction 
#@Henry Lieberman
#t2009
#c408
#index841
#id865
#%
#!

#*term weighting approaches automatic text retrieval experimental evidence past years indicates systems based assignment weighted single terms produce retrieval results superior text representations results depend choice effective term weighting systems paper automatic term weighting provides baseline single term indexing models content analysis procedures compared 
#@Gerard Salton, Chris Buckley
#t2009
#c409
#index842
#id866
#%
#!

#*connectivity server fast access linkage information web built server provides linkage information pages indexed search engine basic operation server query consisting set list pages point pages list pages pages generally server produce entire graph theory sense given distance include information links exist pages information retrieved directly search engines engines optimized purpose process constructing given set pages slow contrast prototype server needs result far built applications use connectivity server direct interface fast navigation web relation visualization tool given set pages numerous applications ranking visualization classification 
#@Krishna Bharat, AZ Broder, Monika Rauch Henzinger, Puneet Kumar, Suresh Venkatasubramanian
#t2009
#c46
#index843
#id867
#%
#!

#*search engines world wide web comparative study evaluation methodology web search engines compared evaluated terms search capabilities boolean logic field search word phrase search retrieval precision response time using sample queries real reference questions recall evaluation criterion information retrieval study assume relevant items particular query changing web authors study search retrieval performance largest coverage web resources web search engines examined result research proposed methodology evaluating web search engines current study 
#@H Chu, M Rosenthal
#t2009
#c410
#index844
#id868
#%
#!

#*interactive query expansion online retrieval appropriate offer users improve initial queries form short lists suggested terms feedback nearest term variants original query terms approach series experiments run using test collection discover techniques select terms lists effective retrieval results significant improvement expected approach query expansion 
#@Donna Harman
#t2009
#c35
#index845
#id869
#%
#!

#*world wide web exploratory analysis structure exploratory study examines explosive growth world wide web based analysis web pages collected web use search engine analysis set science related www sites statistical characteristics web documents hypertext links examined characteristics highly web documents 
#@Ray R. Larson
#t2009
#c411
#index846
#id870
#%
#!

#*interfaces end user information discusses essential features interfaces support end user information include engineering task models task analysis problem solving nature information examples systems end users including online public access catalogs hypertext help systems suggested research directions 
#@Gary Marchionini
#t2009
#c39
#index847
#id871
#%
#!

#*extracting structures web current implementation world wide web explicit structure strong typing closed hypertext systems property directly explosive web difficult problem identifying structures large hypertext collections reduced structures form basis navigation complex hypertext systems 
#@Peter Pirolli, James E. Pitkow, Ramana Rao
#t2009
#c412
#index848
#id872
#%
#!

#*learning block importance models web pages previous works web page partitioned multiple segments blocks usually importance blocks page noisy blocks pages facilitate web mining search works uniform approach model presented measure importance different portions web pages user study people consistent view importance blocks web pages paper investigate model automatically assign importance values blocks web page define block importance estimation learning problem use vision based page segmentation algorithm partition web page semantic blocks hierarchy structure spatial features position size content features number images links extracted construct feature vector block based features learning algorithms svm neural network applied various block importance models experiments best model achieve performance accuracy quite close 
#@Ruihua Song, Haifeng Liu, Ji-Rong Wen, Wei-Ying Ma
#t2009
#c38
#index849
#id873
#%850,851,852,853,854,855,856,857,858
#!

#*detection data mining applications propose detection problem suggest practical solution based frequent item sets use web principles characterize assumptions hypertext information retrieval ir data mining systems major source principles basic pure implementations simple search algorithms coupled detection elimination increases precision levels recall 
#@Ziv Bar-Yossef, S. Rajagopalan
#t2009
#c38
#index850
#id874
#%
#!

#*function based object model adaptation content understanding crucial issue adaptation paper present function based object model fom attempts understand authors identifying object function instead semantic understanding object serves certain functions basic specific function authors purpose object based consideration proposed fom model understanding fom includes parts basic fom based basic functional properties object specific fom based category object automatic approach detect functional properties category object presented fom generation level adaptation rules general rules specific rules based fom combined practical adaptation web content adaptation wireless application protocol developed application example proposed model experiments shown results 
#@Jinlin Chen, Baoyao Zhou, Jin Shi, HongJiang Zhang, Qiu Fengwu
#t2009
#c38
#index851
#id875
#%
#!

#*statistical properties systems linear applications pattern recognition paper separating nonlinear decision direct application theorem classical combinatorial geometry shown family having natural separating pattern vectors extending unifying results pattern separating applying ideas vertices binary cube yields bounds number general boolean functions variables 
#@TM Cover
#t2009
#c413
#index852
#id876
#%
#!

#*error correcting output codes general method improving inductive learning programs learning problems involve finding definition unknown function range discrete set containing values classes definition acquired studying large collections training examples form existing approaches problem include direct application algorithms decision tree algorithms application binary concept learning algorithms learn individual binary functions classes application binary concept learning algorithms distributed output codes employed paper compares approaches new technique error correcting codes employed distributed output representation output representations improve performance task speech recognition task results demonstrate error correcting output codes provide general purpose method improving performance inductive learning programs problems 
#@Thomas G. Dietterich, Ghulum Bakiri
#t2009
#c103
#index853
#id877
#%
#!

#*discovering content blocks web documents paper propose new approach discover contents set documents web pages web site partitions page content blocks according html web page based occurrence features terms set pages entropy value feature according entropy value feature content block entropy value block defined analyzing 
#@Shian-Hua Lin, Jan-Ming Ho
#t2009
#c3
#index854
#id878
#%
#!

#*automatic browsing large mobile devices increasingly common popular mobile communications limitation mobile devices need develop new technologies facilitate browsing large small paper propose novel approach able navigation large minimal user interaction mobile devices image attention model employed illustrate information structure image optimal image browsing path based image attention model human browsing experimental evaluations proposed mechanism indicate approach effective way large images small 
#@Hao Liu, Xing Xie, Wei-Ying Ma, HongJiang Zhang
#t2009
#c414
#index855
#id879
#%
#!

#*web page web mining feature weighting unlike conventional data text web pages typically contain large information main contents pages navigation information web page noise web pages harm web mining clustering classification paper propose novel feature weighting technique deal web page noise enhance web mining method compressed structure tree capture common structure comparable blocks set web pages uses information based measure evaluate importance node compressed structure tree based tree node importance values method assigns weight word feature content block resulting weights used web mining evaluated proposed technique web mining tasks web page clustering web page classification experimental results weighting method able dramatically improve mining results 
#@Lan Yi, Bing Liu
#t2009
#c238
#index856
#id880
#%
#!

#*noisy information web pages data mining commercial web page typically contains information blocks main content blocks usually blocks navigation privacy advertisements business purposes easy user access blocks main content blocks page noisy blocks information contained noisy blocks harm web data mining great importance paper propose noise elimination technique based following observation given web site noisy blocks usually share common contents main content blocks pages diverse actual contents based observation propose tree structure called style tree capture common actual contents pages given web site sampling pages site style tree built site site style tree introduce information based measure determine parts represent parts represent main contents site employed detect web page site mapping page proposed technique evaluated data mining tasks web page clustering classification experimental results noise elimination technique able improve mining results significantly 
#@Lan Yi, Bing Liu, Xiaoli Li
#t2009
#c3
#index857
#id881
#%
#!

#*improving pseudo relevance feedback web information retrieval using web page segmentation contrast traditional document retrieval web page good information unit search contains multiple topics information navigation interaction page paper propose vision based page segmentation algorithm detect semantic content structure web page compared simple based segmentation method page segmentation scheme utilizes useful visual obtain better partition page semantic level using algorithm selection query expansion terms pseudo relevance feedback web information retrieval achieve performance improvement web track dataset 
#@Shipeng Yu, Deng Cai, Ji-Rong Wen, Wei-Ying Ma
#t2009
#c38
#index858
#id882
#%
#!

#*link unified link analysis framework multi type data objects web link analysis proven significant quality based web search existing links classified categories intra type links web represent relationship data objects homogeneous data type web pages inter type links user browsing log represent relationship data objects different data types users web pages unfortunately link analysis research type link paper propose unified link analysis framework called link inter intra type link structure multiple type inter related data objects order objects data type time pagerank hits algorithms shown special cases unified link analysis framework experiments framework makes use user data web pages extracted log proposed algorithm improve search effectiveness hits algorithms respectively 
#@Wensi Xi, Benyu Zhang, Zheng Chen, Yizhou Lu, Shuicheng Yan, Wei-Ying Ma, Edward A. Fox
#t2009
#c38
#index859
#id883
#%836,837,860,861,862,863,864,865,866,867,170
#!

#*mining web link structure web hypertext approximately million pages grow roughly million pages page variation data scale taken set web pages unifying structure shows far style content variation seen traditional text document collections level complexity makes database management information retrieval solution date index based search engines web primary tool users search information engines build indices let quickly retrieve set web pages containing given word users make effective use engines tasks solved searching constrained keywords phrases search engines wide range equally important tasks particular topic typically contain million relevant web pages pages search engine select correct ones value user 
#@Soumen Chakrabarti, Byron Dom, Ravi Kumar, Prabhakar Raghavan, S. Rajagopalan, Andrew Tomkins, D. Gibson, Jon M. Kleinberg
#t2009
#c188
#index860
#id884
#%
#!

#*learning identify authoritative documents model document citation learns identify set linked documents pages retrieved world wide web papers retrieved research paper unlike popular hits algorithm statistical assumptions model provides probabilistic estimates clear semantics general identified authoritative documents correspond better human 
#@David Cohn, Huan Chang
#t2009
#c22
#index861
#id885
#%
#!

#*overview trec web track trec web track away non web relevance ranking tasks million page topic task involved finding pages relevant characteristics make desirable list key pages named page task variant years finding task task particular page years task page need home page 
#@Nick Craswell, David Hawking
#t2009
#c52
#index862
#id886
#%
#!

#*effective site finding using link information link based ranking methods described literature applied commercial web search engines according recent trec experiments better traditional content based methods conduct different type experiment task main entry point specific web site experiments ranking based link text effective ranking based document content methods used obtained results using sets queries million document set set million document set site finding effectiveness explain search engines adopted link methods rich new area effectiveness improvement traditional methods 
#@Nick Craswell, David Hawking, S.E. Robertson
#t2009
#c35
#index863
#id887
#%
#!

#*text link analysis paper presents simple idea relationships terms documents generate representation aspects web link analysis text analysis techniques information retrieval paper shows path unified representation demonstrates use eigenvector web link analysis simple example 
#@Brian D. Davison
#t2009
#c35
#index864
#id888
#%
#!

#*citation analysis tool journal evaluation communications network play role scientific technical information little published study references years issues journal pieces network work map journal network date studies network components limited number areas scientific study time authors able consider purpose insight energy practical difficulty manually necessary data 
#@Eugene Garfield
#t2009
#c252
#index865
#id889
#%
#!

#*graph theory practice world wide web answer web world wide according university web question geometric distance concept mathematics called graph theory web place place hypertext links makes sense define distance steps links question select web pages random links separate average million pages web long paths et al know just calculation interesting style technology graph theory just years apply graph theoretical methods structure world wide web course just years web exist large objects graph theoretical analysis time graph theory graph practice graph engineering 
#@B Hayes
#t2009
#c415
#index866
#id890
#%
#!

#*new index derived analysis purpose evaluating manner free popularity procedures paper presents new method computation takes choose necessary introduce connection concept influence 
#@L. Katz
#t2009
#c416
#index867
#id891
#%
#!

#*object level ranking order web objects contrast current web search methods essentially document level ranking retrieval exploring new paradigm enable web search object level collect web information objects relevant application domain rank objects terms relevance popularity answer user queries traditional pagerank model longer valid object popularity calculation heterogeneous relationships objects paper introduces domain independent object level link analysis model rank objects domain assign popularity propagation factor type object relationship study different popularity propagation factors heterogeneous relationships affect popularity ranking propose approaches automatically decide factors experiments using million papers experimental results achieve better ranking results applying pagerank object graph 
#@Zaiqing Nie, Yuanzhi Zhang, Ji-Rong Wen, Wei-Ying Ma
#t2009
#c38
#index868
#id892
#%869,870,871,872,873,874,170,875,876,849
#!

#*wrapper generation semi structured internet sources current information world wide web www wealth information different subjects available line numerous sources contain information classified semi structured present way access information browsing individual pages query web documents database like fashion based underlying structure provide database like querying semi structured www sources building sources present approach semi automatically generating key idea exploit information pages source underlying structure page structure wrapper querying source possibly integrating sources demonstrate able build number internet sources different domains using implemented wrapper generation 
#@Naveen Ashish, Craig A. Knoblock
#t2009
#c328
#index869
#id893
#%
#!

#*authority based keyword queries databases using applies authority based ranking keyword search databases modeled labeled graphs conceptually authority nodes objects containing keywords objects according semantic connections node ranked according authority respect particular keywords adjust weight global importance weight keyword query importance result containing keywords versus 
#@Andrey Balmin, Vagelis Hristidis, Yannis Papakonstantinou
#t2009
#c417
#index870
#id894
#%
#!

#*block level link analysis link analysis shown great potential improving web search pagerank hits popular algorithms existing link analysis algorithms treat web page single node web graph cases web page contains multiple semantics web page considered node paper web page partitioned blocks using vision based page segmentation algorithm extracting page block block page relationships link structure page layout analysis construct semantic graph www node represents single semantic topic graph better semantic structure web based block level link analysis proposed new algorithms block level pagerank block level hits study extensively using web data 
#@Deng Cai, Xiaofei He, Ji-Rong Wen, Wei-Ying Ma
#t2009
#c35
#index871
#id895
#%
#!

#*ranked keyword search xml documents consider problem efficiently producing ranked results keyword search queries hyperlinked xml documents evaluating keyword search queries hierarchical xml documents conceptually html documents introduces new challenges xml keyword search queries return entire documents return nested xml elements contain desired keywords second nested structure xml notion ranking 
#@Lin Guo, Feng Shao, Chavdar Botev, Jayavel Shanmugasundaram
#t2009
#c10
#index872
#id896
#%
#!

#*discovering complex matchings web query interfaces correlation mining approach enable information integration matching critical step discovering semantic attributes heterogeneous sources complex matchings common far complex search space existing techniques focus simple matchings paper takes conceptually novel approach matching correlation mining task matching web query interfaces integrate databases internet web query interfaces generally form complex matchings attribute groups books domain observe occurrences patterns query interfaces reveal complex semantic relationships grouping attributes tend present query interfaces correlated contrast attributes correlated occur insight enables discover complex matchings correlation mining approach particular develop dcm framework consists data dual mining positive negative correlations finally matching selection unlike previous correlation mining algorithms focus finding strong positive correlations algorithm positive negative correlations especially negative correlations special importance matching leads introduction new correlation measure measure proposed previous work evaluate approach extensively results good accuracy discovering complex matchings 
#@Bin He, Kevin Chen-Chuan Chang, Jiawei Han
#t2009
#c3
#index873
#id897
#%
#!

#*optimization useful connection statistical behavior systems finite multivariate combinatorial optimization finding minimum given function depending parameters detailed provides framework optimization properties large complex systems connection statistical new information provides perspective traditional optimization problems methods 
#@S. Kirkpatrick, C. D. Gelatt, M. P. Vecchi
#t2009
#c252
#index874
#id898
#%
#!

#*wrapper induction information extraction internet presents numerous sources useful information product catalogs stock recently systems built automatically gather information user resources usually use people relevant content embedded html pages extracting content difficult used purpose wrapper procedure 
#@Nicholas Kushmerick, Daniel S. Weld, Robert B. Doorenbos
#t2009
#c408
#index875
#id899
#%
#!

#*mining data records web pages large information web contained structured objects data records data records important present essential information host pages lists products services useful data records order extract information provide value added services existing approaches solving problem include manual approach supervised learning automatic 
#@Bing Liu, Robert L. Grossman, Yanhong Zhai
#t2009
#c3
#index876
#id900
#%
#!

#*automatic based web usage mining speed business transactions carried web key rapid growth commerce ability track user browsing behavior individual end customer possible product messages individual customers massive scale phenomenon course type 
#@Bamshad Mobasher, Robert Cooley, Jaideep Srivastava
#t2009
#c7
#index877
#id901
#%
#!

#*personal design implementation growing availability information sources especially non homogeneous distributed sources like world wide web growing tools help making good selection information interested 
#@Dunja Mladenic
#t2009
#c119
#index878
#id902
#%
#!

#*identifying interesting web sites software agent learns rate pages world wide web www pages user user rates explored pages point scale learns user profile analyzing information page user profile used ways used suggest links user interested exploring second used construct query pages user compare different algorithms machine learning information retrieval task naive bayesian classifier offers advantages learning algorithms task furthermore initial portion web page sufficient making predictions substantially reducing network required make predictions 
#@M Pazzani, J Muramatsu, D Billsus
#t2009
#c103
#index879
#id903
#%
#!

#*feature weighting instance selection collaborative filtering collaborative filtering uses database consumers preferences make personal product recommendations achieving widespread success commerce paper present feature weighting methods improve accuracy collaborative filtering algorithms furthermore propose reduce training data set selecting highly relevant instances evaluate various methods known eachmovie data set experimental results mutual information achieves largest accuracy gain feature weighting methods interesting fact data reduction method achieves improvement accuracy collaborative filtering algorithm factor 
#@Kai Yu, Zhong Wen, Xiaowei Xu, Martin Ester
#t2009
#c418
#index880
#id904
#%
#!

#*instance selection techniques memory based collaborative filtering collaborative filtering cf important data mining technique make personalized recommendations books web pages movies popular algorithm memory based collaborative filtering users preference based similar