AI helps you reading Science

AI generates interpretation videos

AI extracts and analyses the key points of the paper to generate videos automatically


pub
Go Generating

AI Traceability

AI parses the academic lineage of this thesis


Master Reading Tree
Generate MRT

AI Insight

AI extracts a summary of this paper


Weibo:
In this work we have presented a new Web Usage Mining recommender system, called SUGGEST 3.0, that is able to dynamically generated personalized content in order to make easier the Web user navigation

An Online Recommender System for Large Web Sites

Web Intelligence, pp.199-205, (2004)

Cited by: 94|Views8
EI WOS

Abstract

In this paper we propose a WUM recommender system, called SUGGEST 3.0, that dynamically generates links to pages that have not yet been visited by a user and might be of his potential interest. Differently from the recommender systems proposed so far, SUGGEST 3.0 does not make use of any off-line component, and is able to manage Web sites...More

Code:

Data:

0
Introduction
  • The continuous and rapid growth of the Web has led to the development of new methods and tools in the Web recommender or personalization domain [4], [2].
  • The authors present a WUM system, called SUGGEST 3.0, which is designed to dynamically generated personalized content of potential interest for users of a Web Site.
  • The adoption of a LRU-based algorithm to manage the knowledge base permits them to use SUGGEST 3.0 on large Web sites made up of pages dynamically generated.
Highlights
  • The continuous and rapid growth of the Web has led to the development of new methods and tools in the Web recommender or personalization domain [4], [2]
  • It is able to update incrementally and automatically the knowledge base obtained from historical usage data and to generate a list of page links
  • In order to manage large Web Sites that may require an adjacency matrix that exceeds the maximum available memory, we adopted a LRU-based strategy to store in M only those pages that have been recently accessed by some users
  • In this work we have presented a new Web Usage Mining (WUM) recommender system, called SUGGEST 3.0, that is able to dynamically generated personalized content in order to make easier the Web user navigation
  • The proposed system is composed by a single component, tightly integrated with the Apache Web server. It is based on an incremental procedure, that is able to update incrementally and automatically the knowledge base obtained from historical usage data and to generate a list of links to pages of potentially interest for the user
  • Experimental results show that SUGGEST 3.0 is able to generate valid suggestions with a limited overhead on the Web server
Results
  • SUGGEST 3.0 works as follows: once a new request arrives at the server, the URL requested and the session to which the user belongs are identified, the underlying knowledge base is updated, and a list of suggestions is appended to the requested page.
  • According to the current session characteristics, it updates the knowledge base and generates the suggestions to be presented to the user.
  • The use of the cookie mechanism remove a drawback present in SUGGEST 2.0 that identified user sessions by applying a heuristic based on the IP address and time-stamp.
  • In order to manage large Web Sites that may require an adjacency matrix that exceeds the maximum available memory, the authors adopted a LRU-based strategy to store in M only those pages that have been recently accessed by some users.
  • For each test the authors generated requests to an Apache server running SUGGEST 3.0 and recorded the suggestions generated for every navigation session contained within the access log file considered.
  • Since SUGGEST 3.0 is able to manage Web sites with dynamic pages it will obtain the performance similar to those obtainable by SUGGEST 2.0 on the same site when the number of accessed pages reaches its maximum value.
  • In this work the authors have presented a new WUM recommender system, called SUGGEST 3.0, that is able to dynamically generated personalized content in order to make easier the Web user navigation.
  • It is based on an incremental procedure, that is able to update incrementally and automatically the knowledge base obtained from historical usage data and to generate a list of links to pages of potentially interest for the user.
Conclusion
  • The adoption of a LRU-based algorithm to manage the knowledge base, permits them to use the system on Web sites that exploit pages dynamically generated (i.e. Web site made up of a not fixed number of pages).
  • To this end the authors think to extend the classical PageRank algorithm [10] to evaluate the page relevance using both the information about the site linkage structure, and the information extracted from the historical Web usage data.
  • The exploitation of the suggestion can lead to reduce the average session length improving the performance of the Web server
Summary
  • The continuous and rapid growth of the Web has led to the development of new methods and tools in the Web recommender or personalization domain [4], [2].
  • The authors present a WUM system, called SUGGEST 3.0, which is designed to dynamically generated personalized content of potential interest for users of a Web Site.
  • The adoption of a LRU-based algorithm to manage the knowledge base permits them to use SUGGEST 3.0 on large Web sites made up of pages dynamically generated.
  • SUGGEST 3.0 works as follows: once a new request arrives at the server, the URL requested and the session to which the user belongs are identified, the underlying knowledge base is updated, and a list of suggestions is appended to the requested page.
  • According to the current session characteristics, it updates the knowledge base and generates the suggestions to be presented to the user.
  • The use of the cookie mechanism remove a drawback present in SUGGEST 2.0 that identified user sessions by applying a heuristic based on the IP address and time-stamp.
  • In order to manage large Web Sites that may require an adjacency matrix that exceeds the maximum available memory, the authors adopted a LRU-based strategy to store in M only those pages that have been recently accessed by some users.
  • For each test the authors generated requests to an Apache server running SUGGEST 3.0 and recorded the suggestions generated for every navigation session contained within the access log file considered.
  • Since SUGGEST 3.0 is able to manage Web sites with dynamic pages it will obtain the performance similar to those obtainable by SUGGEST 2.0 on the same site when the number of accessed pages reaches its maximum value.
  • In this work the authors have presented a new WUM recommender system, called SUGGEST 3.0, that is able to dynamically generated personalized content in order to make easier the Web user navigation.
  • It is based on an incremental procedure, that is able to update incrementally and automatically the knowledge base obtained from historical usage data and to generate a list of links to pages of potentially interest for the user.
  • The adoption of a LRU-based algorithm to manage the knowledge base, permits them to use the system on Web sites that exploit pages dynamically generated (i.e. Web site made up of a not fixed number of pages).
  • To this end the authors think to extend the classical PageRank algorithm [10] to evaluate the page relevance using both the information about the site linkage structure, and the information extracted from the historical Web usage data.
  • The exploitation of the suggestion can lead to reduce the average session length improving the performance of the Web server
Tables
  • Table1: Access log files used to measure the suggestions quality
Download tables as Excel
Related work
  • In the past, several WUM projects have been proposed to foresee users preference and their navigation behavior. In the following we review some of the most significant WUM projects that can be compared with our system.

    Analog [16] is one of the first WUM systems. It is structured according to an off-line and an online component. The off-line component builds session clusters by analyzing past users activity recorded in server log files. Then the online component builds active user sessions which are then classified according to the generated model. The classification allows to identify pages related to the ones in the active session and to return the requested page with a list of suggestions. The geometrical approach used for clustering is affected by several limitations, related to scalability and to the effectiveness of the results found. Nevertheless, the architectural solution introduced was maintained in several other more recent projects.
Funding
  • 1This work was funded by the Italian Ministry of Education, University and Research (MIUR) as part of the National Project Legge 449/97, 1999, settore Societadell’Informazione: Technologies and Services for Enhanced Contents Delivery (2002-2004)
Study subjects and analysis
cases: 3
Apache Suggest 2.0 Suggest 3.0. In all the three cases, as the number of concurrent requests increases the request response time increases proportionally. It is due to the mutual exclusive accesses to shared memory areas by the Apache processes

Reference
  • R. Baraglia and P. Palmerini. Suggest: A web usage mining system. In Proc. of IEEE Int’l Conf. on Information Technology: Coding and Computing, April 2002.
    Google ScholarLocate open access versionFindings
  • M. Deshpande and G. Karypis. Item-based top-n recommendation algorithms. ACM Trans. on Information Systems, 22(1):143–177, January 2004.
    Google ScholarLocate open access versionFindings
  • R. Kosala and H. Blockeel. Web mining research: A survey. ACM SIGKDD, 2(1):1–15, July 2000.
    Google ScholarLocate open access versionFindings
  • E. Magdalini and M. Vazirgiannis. Web mining for web personalization. ACM Trans. on Internet Technology, 3(1):1– 27, February 2003.
    Google ScholarLocate open access versionFindings
  • B. Mobasher, R. Cooley, and J. Srivastava. Automatic personalization based on web usage mining. Communications of the ACM, 43(8):142–151, august 2000.
    Google ScholarLocate open access versionFindings
  • B. Mobasher, N. Jain, E.-H. S. Han, and J. Srivastava. Web mining: Pattern discovery from world wide web transactions. TR 96-050, University of Minnesota, 1996.
    Google ScholarFindings
  • M. D. Mulvenna, S. S. Anand, and A. G. Buchener. Personalization on the net using web mining. Communication of ACM, 43(8), 2000.
    Google ScholarLocate open access versionFindings
  • M. Nakagawa and B. Mobasher. A hybrid web personalization model based on site connectivity. In Proc. of WebKDD, pages 59–70, 2003.
    Google ScholarLocate open access versionFindings
  • O. Nasraoui and C. Petenes. Combining web usage mining and fuzzy inference for website personalization. In Proc. of WebKDD, pages 37–46, 2003.
    Google ScholarLocate open access versionFindings
  • L. Page, S. Brin, R. Motwani, and T. Winograd. The Pagerank Citation Ranking: Bringing Order to the Web. Technical report, Stanford University, 1998.
    Google ScholarFindings
  • M. Perkowitz and O. Etzioni. Adaptive web sites: Conceptual cluster mining. In Int’l Joint Conf. on AI, pages 264– 269, 1999.
    Google ScholarLocate open access versionFindings
  • J. G. Siek, L. Lee, and A. Lumsdaine. Boost Graph Library, The: User Guide and Reference Manual. Addison Wesley Professional, 2001.
    Google ScholarFindings
  • F. Silvestri, R. Baraglia, P. Palmerini, and S. M. On-line generation of suggestions for web users. In Proc. of IEEE Int’l Conf. on Information Technology: Coding and Computing, April 2004.
    Google ScholarLocate open access versionFindings
  • R. Thau. Design considerations for the Apache Server API. Computer Networks and ISDN Systems, 28(7–11):1113– 1122, 1996.
    Google ScholarLocate open access versionFindings
  • K.-l. Wu, P. S. Yu, and A. Ballman. Speedtracer: A web usage mining and analysis tool. IBM Systems Journal, 37(1), 1998.
    Google ScholarLocate open access versionFindings
  • T. W. Yan, M. Jacobsen, H. Garcia-Molina, and D. Umeshwar. From user access patterns to dynamic hypertext linking. Fifth International World Wide Web Conference, May 1996.
    Google ScholarFindings
Your rating :
0

 

Tags
Comments
小科