AI帮你理解科学

AI 生成解读视频

AI抽取解析论文重点内容自动生成视频


pub
生成解读视频

AI 溯源

AI解析本论文相关学术脉络


Master Reading Tree
生成 溯源树

AI 精读

AI抽取本论文的概要总结


微博一下
We have presented a solution to the privacy-preserving collaborative filtering problem using the randomized perturbation scheme

Privacy-Preserving Collaborative Filtering.

IEEE International Conference on Data Mining, no. 4 (2014): 9-35

引用84|浏览66
EI
下载 PDF 全文
引用
微博一下

摘要

Collaborative Filtering (CF) techniques are becoming increasingly popular with the evolution of the Internet. E-commerce sites use CF systems to suggest products to customers based on like-minded customers’ preferences. People use CF systems to cope with information overload. To conduct collaborative filtering, data from customers are nee...更多

代码

数据

0
简介
  • With the amount of the information available for individuals growing steadily, information overload has become a major problem for users.
  • To make the information to serve users better, information filtering and recommendation schemes become more and more important.
  • With the number of users accessing the Internet growing, CF techniques are becoming increasingly popular as part of online shopping sites.
  • These sites incorporate recommendation systems that suggest products to users based on products that like-minded users have ordered before, or indicate as interesting.
  • Users can get recommendations about many of their daily activities, including restaurants, bars, movies, books, news, music CDs, interesting sights to see, and things to do in a city with the help of collaborative filtering
重点内容
  • With the amount of the information available for individuals growing steadily, information overload has become a major problem for users
  • We propose a randomized perturbation technique to protect users’ privacy while still producing accurate recommendations
  • Our results show that the Collaborative filtering systems using the randomized perturbation techniques provide accurate recommendations while preserving the users’ privacy
  • While Canny’s work focuses on the peer-to-peer framework, in which users actively participate in the collaborative filtering process, our work focuses on another framework, in which users send their data to a central place and they do not participate in the Collaborative filtering process; only the central place needs to conduct the Collaborative filtering
  • We have presented a solution to the privacy-preserving collaborative filtering problem using the randomized perturbation scheme
  • We believe that accuracy of our scheme can be further improved if more aggregate information is disclosed along with the disguised data, especially those aggregate information whose disclosure does not compromise much of users’ privacy
方法
  • The outline of the procedure is described in the following: 1.
  • The authors randomly divided the dataset into a training set (900 users) and a testing set (43 users).
  • For each active user selected randomly from the testing dataset, the authors randomly select an item and use the randomized-perturbation-based scheme and the original algorithm, respectively, to predict the ratings on this item for this active user.
  • The authors calculate the difference of these two ratings
  • The authors run this prediction procedure for 100 times and calculate the mean absolute error and standard deviation of the errors
结果
  • 4.1 Datasets

    The authors use Jester and MovieLens datasets in the experiments to evaluate the accuracy of the randomizedperturbation-based collaborative filtering scheme.
  • Some users end up reading and rating all the jokes, so Jester is much more dense than the other datasets the authors used.
  • To evaluate the proposed schemes, the authors have conducted several experiments; the authors compared the prediction results from randomized data with the results from the original data.
  • Fig. 3, Fig. 4, and Fig. 5 depict the results on three different datasets
  • For both MovieLens datasets (MovieLens public data and MovieLens million data), when the authors choose γ = 95% and the fixed-α scheme, the mean absolute error in the experiments is below 0.29.
结论
  • Conclusion and Future Work

    The authors have presented a solution to the privacy-preserving collaborative filtering problem using the randomized perturbation scheme.
  • The authors believe that accuracy of the scheme can be further improved if more aggregate information is disclosed along with the disguised data, especially those aggregate information whose disclosure does not compromise much of users’ privacy.
  • These types of information include mean, standard deviation, distribution, true data in a permuted order, etc.
  • The authors will study how these kinds of aggregate data disclosure affects the accuracy and the privacy
表格
  • Table1: Examples
Download tables as Excel
相关工作
  • 2.1 Privacy-Preserving Collaborative Filtering

    Canny proposes two schemes for privacy-preserving collaborative filtering [4, 5]. In these schemes, users control all of their own private data; a community of users can compute a public “aggregate” of their data without disclosing individual users’ data. The aggregate allows personalized recommendations to be computed by members of the community, or by outsiders. Canny’s method reduces the collaborative filtering task to an iterative calculation of the aggregate requiring only addition of vectors of user data. Canny then uses homomorphic encryption to allow sums of encrypted vectors to be computed and decrypted without exposing individual data. His schemes are based on distributed computation of a certain aggregate of all users’ data. The aggregate is treated as public data. Each user constructs the aggregate and uses local computation to get personalized recommendations. Canny’s schemes can be implemented with untrusted servers, or with additional infrastructures, as a fully peer-to-peer (P2P) system. The P2P architecture allows users to create and maintain their own recommender groups themselves.
引用论文
  • Anonymizer.com: http://www.anonymizer.com.
    Findings
  • R. Agrawal and R. Srikant. Privacy-preserving data mining. In Proceedings of the 2000 ACM SIGMOD on Management of Data, pages 439–450, Dallas, TX USA, May 15 - 18 2000.
    Google ScholarLocate open access versionFindings
  • J. Breese, D. Heckerman, and C. Kadie. Empirical analysis of predictive algorithms for collaborative filtering. In Proceedings of the Fourteenth Conference on Uncertainty in Artificial Intelligence, pages 43–52, Madison,WI, July 1998.
    Google ScholarLocate open access versionFindings
  • J. Canny. Collaborative filtering with privacy. In IEEE Symposium on Security and Privacy, pages 45–57, Oakland, CA, May 2002.
    Google ScholarLocate open access versionFindings
  • J. Canny. Collaborative filtering with privacy via factor analysis. In Proceedings of the 25th annual international ACM SIGIR conference on Research and development in information retrieval, pages 238–245, Tampere, Finland, August 2002.
    Google ScholarLocate open access versionFindings
  • L. F. Cranor, J. Reagle, and M. S. Ackerman. Beyond concern: Understanding net users’ attitudes about online privacy. Technical report, AT&T Labs-Research, April 1999. Available from http://www.research.att.com/library/trs/TRs/99/99.4.3/report.htm.
    Findings
  • D. Goldberg, D. Nichols, B. M. Oki, and D. Terry. Using collaborative filtering to weave an information tapestry. 35(12):61–70, 1992.
    Google ScholarFindings
  • K. Goldberg, T. Roeder, D. Gupta, and C. Perkins. Eigentaste: A constant time collaborative filtering algorithm. Information Retrieval, 4(2):133–151, 2001.
    Google ScholarLocate open access versionFindings
  • D. Gupta, M. Digiovanni, H. Narita, and K. Goldberg. Jester 2.0: A new linear-time collaborative filtering algorithm applied to jokes. In Workshop on Recommender Systems Algorithms and Evaluation, 22nd International Conference on Research and Development in Information Retrieval, Berkeley, CA.
    Google ScholarLocate open access versionFindings
  • J. Herlocker, J. Konstan, A. Borchers, and J. Riedl. An algorithmic framework for performing collaborative filtering. In Proceedings of the 1999 Conference on Research and Development in Information Retrieval, August 1999.
    Google ScholarLocate open access versionFindings
  • W. Hill, L. Stead, M. Rosenstein, and G. Furnas. Recommending and evaluating choices in a virtual community of use. In Proceedings of ACM CHI’95 Conference on Human Factors in Computing Systems, pages 194–201, 1995.
    Google ScholarLocate open access versionFindings
  • Joseph A. Konstan, Bradley N. Miller, David Maltz, Jonathan L. Herlocker, Lee R. Gordon, and John Riedl. Grouplens: Applying collaborative filtering to usenet news. In Communications of the ACM, pages 77–87, March 1997.
    Google ScholarLocate open access versionFindings
  • David M. Pennock, Eric Horvitz, Steve Lawrence, and C. Lee Giles. Collaborative filtering by personality diagnosis: A hybrid memory and model based approach. In Proceedings of the Sixteenth Conference on Uncertainty in Artificial Intelligence (UAI-2000), pages 473–480, Morgan Kaufmann, San Francisco, 2000.
    Google ScholarLocate open access versionFindings
  • M. K. Reiter and A. D. Rubin. Crowds: anonymity for web transaction. ACM Transactions on Information and System Security, 1(1):Pages 66–92, 1998.
    Google ScholarLocate open access versionFindings
  • P. Resnick, N. Iacovou, M. Suchak, P. Bergstrom, and J. Riedl. Grouplens: An open architecture for collaborative filtering of netnews. In Proceedings of the ACM Conference on Computer Supported Cooperative Work, pages 175–186, 1994.
    Google ScholarLocate open access versionFindings
  • B. M. Sarwar, G. Karypis, J. A. Konstan, and J. T. Riedl. Application of dimensionality reduction in recommender system-a case study. In ACM WebKDD 2000 Web Mining for E-commerce Workshop, 2000.
    Google ScholarLocate open access versionFindings
  • Upendra Shardanand and Patti Maes. Social information filtering: Algorithms for automating ”word of mouth”. In Proceedings of ACM CHI’95 Conference on Human Factors in Computing Systems, pages 210–217, 1995.
    Google ScholarLocate open access versionFindings
  • P. F. Syverson, D. M. Goldschlag, and M. G. Reed. Anonymous connections and onion routing. In Proceedings of 1997 IEEE Symposium on Security and Privacy, Oakland, California, USA, May 5-7 1997.
    Google ScholarLocate open access versionFindings
  • A. F Westin. Freebies and privacy. Technical report, Opinion Research Corporation, July 1999. Availabe from http://www.privacyexchange.org/iss/surveys/sr990714.html.
    Findings
0
您的评分 :

暂无评分

标签
评论
avatar
数据免责声明
页面数据均来自互联网公开来源、合作出版商和通过AI技术自动分析结果,我们不对页面数据的有效性、准确性、正确性、可靠性、完整性和及时性做出任何承诺和保证。若有疑问,可以通过电子邮件方式联系我们:report@aminer.cn