Item-based collaborative filtering recommendation algorithms

    WWW, pp. 285-295, 2001.

    Cited by: 8121|Bibtex|Views110|Links
    EI
    Keywords:
    recommender systemcollaborative filteringresiliencyscalabilityk nearest neighborMore(7+)
    Wei bo:
    New technologies are needed that can dramatically improve the scalability of recommender systems

    Abstract:

    Recommender systems apply knowledge discovery techniques to the problem of making personalized recommendations for information, products or services during a live interaction. These systems, especially the k-nearest neighbor collaborative filtering based ones, are achieving widespread success on the Web. The tremendous growth in the amoun...More

    Code:

    Data:

    Introduction
    • The amount of information in the world is increasing far more quickly than the ability to process it.
    • Collaborative filtering works by building a database of preferences for items by users.
    • A new user, Neo, is matched against the database to discover neighbors, which are other users who have historically had similar taste to Neo. Items that the neighbors like are recommended to Neo, as he will probably like them.
    • There remain important research questions in overcoming two fundamental challenges for collaborative filtering recommender systems
    Highlights
    • The amount of information in the world is increasing far more quickly than our ability to process it
    • Collaborative filtering has been very successful in both research and practice, and in both information filtering applications and E-commerce applications
    • We briefly present some of the research literature related to collaborative filtering, recommender systems, data mining and personalization
    • It can be observed from the results that offsetting the user-average for cosine similarity computation has clear advantage, as the Mean Absolute Error is significantly lower in this case
    • New technologies are needed that can dramatically improve the scalability of recommender systems
    • In this paper we presented and experimentally evaluated a new algorithm for Collaborative Filtering-based recommender systems
    Methods
    • Experiments with neighborhood size

      The size of the neighborhood has significant impact on the prediction quality [12].
    • To determine the sensitivity of this parameter, the authors performed an experiment where the authors varied the number of neighbors to be used and computed MAE.
    • The two methods show different types of sensitivity.
    • The regression-based algorithm shows decrease in prediction quality with increased number of neighbors.
    • Considering both trends the authors select 30 as the optimal choice of neighborhood size
    Results
    • The authors present the experimental results of applying item-based collaborative filtering techniques for generating predictions.
    • The authors implemented the algorithm to compute the neighborhood and used weighted sum algorithm to generate the prediction
    • The authors ran these experiments on the training data and used test set to compute Mean Absolute Error (MAE).
    • Using the training data set the authors precomputed the item similarity using different model sizes and used only the weighted sum prediction generation technique to provide the predictions.
    Conclusion
    • From the experimental evaluation of the item-item collaborative filtering scheme the authors make some important observations.
    • The item-item scheme is capable in addressing the two most important challenges of recommender systems for E-Commerce–quality of prediction and high performance.Recommender systems are a powerful new technology for extracting additional value for a business from its user databases.
    • These systems help users find items they want to buy from a business.
    • New technologies are needed that can dramatically improve the scalability of recommender systems
    Summary
    • Introduction:

      The amount of information in the world is increasing far more quickly than the ability to process it.
    • Collaborative filtering works by building a database of preferences for items by users.
    • A new user, Neo, is matched against the database to discover neighbors, which are other users who have historically had similar taste to Neo. Items that the neighbors like are recommended to Neo, as he will probably like them.
    • There remain important research questions in overcoming two fundamental challenges for collaborative filtering recommender systems
    • Methods:

      Experiments with neighborhood size

      The size of the neighborhood has significant impact on the prediction quality [12].
    • To determine the sensitivity of this parameter, the authors performed an experiment where the authors varied the number of neighbors to be used and computed MAE.
    • The two methods show different types of sensitivity.
    • The regression-based algorithm shows decrease in prediction quality with increased number of neighbors.
    • Considering both trends the authors select 30 as the optimal choice of neighborhood size
    • Results:

      The authors present the experimental results of applying item-based collaborative filtering techniques for generating predictions.
    • The authors implemented the algorithm to compute the neighborhood and used weighted sum algorithm to generate the prediction
    • The authors ran these experiments on the training data and used test set to compute Mean Absolute Error (MAE).
    • Using the training data set the authors precomputed the item similarity using different model sizes and used only the weighted sum prediction generation technique to provide the predictions.
    • Conclusion:

      From the experimental evaluation of the item-item collaborative filtering scheme the authors make some important observations.
    • The item-item scheme is capable in addressing the two most important challenges of recommender systems for E-Commerce–quality of prediction and high performance.Recommender systems are a powerful new technology for extracting additional value for a business from its user databases.
    • These systems help users find items they want to buy from a business.
    • New technologies are needed that can dramatically improve the scalability of recommender systems
    Related work
    • In this section we briefly present some of the research literature related to collaborative filtering, recommender systems, data mining and personalization.

      Tapestry [10] is one of the earliest implementations of collaborative filtering-based recommender systems. This system relied on the explicit opinions of people from a close-knit community, such as an office workgroup. However, recommender system for large communities cannot depend on each person knowing the others. Later, several ratings-based automated recommender systems were developed. The GroupLens research system [19, 16] provides a pseudonymous collaborative filtering solution for Usenet news and movies. Ringo [27] and Video Recommender [14] are email and web-based systems that generate recommendations on music and movies respectively. A special issue of Communications of the ACM [20] presents a number of different recommender systems.
    Funding
    • Funding for this research was provided in part by the National Science Foundation under grants IIS 9613960, IIS 9734442, and IIS 9978717 with additional funding by Net Perceptions Inc
    • This work was also supported by NSF CCR-9972519, EIA-9986042, ACI-9982274 by Army Research Office contract DA/DAAG55-98-1-0441, by the DOE ASCI program and by Army High Performance Computing Research Center contract number DAAH04-95-C-0008. Access to computing facilities was provided by AHPCRC, Minnesota Supercomputer Institute
    Reference
    • Aggarwal, C. C., Wolf, J. L., Wu K., and Yu, P. S. (1999). Horting Hatches an Egg: A New Graph-theoretic Approach to Collaborative Filtering. In Proceedings of the ACM KDD’99 Conference. San Diego, CA. pp. 201212.
      Google ScholarLocate open access versionFindings
    • Basu, C., Hirsh, H., and Cohen, W. (1998). Recommendation as Classification: Using Social and Content-based Information in Recommendation. In Recommender System Workshop ’98. pp. 11-15.
      Google ScholarLocate open access versionFindings
    • Berry, M. W., Dumais, S. T., and O’Brian, G. W. (1995). Using Linear Algebra for Intelligent Information Retrieval. SIAM Review, 37(4), pp. 573-595.
      Google ScholarLocate open access versionFindings
    • Billsus, D., and Pazzani, M. J. (1998). Learning Collaborative Information Filters. In Proceedings of ICML ’98. pp. 46-53.
      Google ScholarLocate open access versionFindings
    • Brachman, R., J., Khabaza, T., Kloesgen, W., Piatetsky-Shapiro, G., and Simoudis, E. 1996. Mining Business Databases. Communications of the ACM, 39(11), pp. 42-48, November.
      Google ScholarLocate open access versionFindings
    • Breese, J. S., Heckerman, D., and Kadie, C. (1998). Empirical Analysis of Predictive Algorithms for Collaborative Filtering. In Proceedings of the 14th Conference on Uncertainty in Artificial Intelligence, pp. 43-52.
      Google ScholarLocate open access versionFindings
    • Cureton, E. E., and D’Agostino, R. B. (1983). Factor Analysis: An Applied Approach. Lawrence Erlbaum associates pubs. Hillsdale, NJ.
      Google ScholarFindings
    • Deerwester, S., Dumais, S. T., Furnas, G. W., Landauer, T. K., and Harshman, R. (1990). Indexing by Latent Semantic Analysis. Journal of the American Society for Information Science, 41(6), pp. 391-407.
      Google ScholarLocate open access versionFindings
    • Fayyad, U. M., Piatetsky-Shapiro, G., Smyth, P., and Uthurusamy, R., Eds. (1996). Advances in Knowledge Discovery and Data Mining. AAAI press/MIT press.
      Google ScholarFindings
    • Goldberg, D., Nichols, D., Oki, B. M., and Terry, D. (1992). Using Collaborative Filtering to Weave an Information Tapestry. Communications of the ACM. December.
      Google ScholarFindings
    • Good, N., Schafer, B., Konstan, J., Borchers, A., Sarwar, B., Herlocker, J., and Riedl, J. (1999). Combining Collaborative Filtering With Personal Agents for Better Recommendations. In Proceedings of the AAAI-’99 conference, pp 439-446.
      Google ScholarLocate open access versionFindings
    • Herlocker, J., Konstan, J., Borchers, A., and Riedl, J. (1999). An Algorithmic Framework for Performing Collaborative Filtering. In Proceedings of ACM SIGIR’99. ACM press.
      Google ScholarLocate open access versionFindings
    • Herlocker, J. (2000). Understanding and Improving Automated Collaborative Filtering Systems. Ph.D. Thesis, Computer Science Dept., University of Minnesota.
      Google ScholarFindings
    • Hill, W., Stead, L., Rosenstein, M., and Furnas, G. (1995). Recommending and Evaluating Choices in a Virtual Community of Use. In Proceedings of CHI ’95.
      Google ScholarLocate open access versionFindings
    • Karypis, G. (2000). Evaluation of Item-Based Top-N Recommendation Algorithms. Technical Report CS-TR-0046, Computer Science Dept., University of Minnesota.
      Google ScholarFindings
    • Konstan, J., Miller, B., Maltz, D., Herlocker, J., Gordon, L., and Riedl, J. (1997). GroupLens: Applying Collaborative Filtering to Usenet News. Communications of the ACM, 40(3), pp. 77-87.
      Google ScholarLocate open access versionFindings
    • Ling, C. X., and Li C. (1998). Data Mining for Direct Marketing: Problems and Solutions. In Proceedings of the 4th International Conference on Knowledge Discovery and Data Mining, pp. 73-79.
      Google ScholarLocate open access versionFindings
    • Peppers, D., and Rogers, M. (1997). The One to One Future: Building Relationships One Customer at a Time. Bantam Doubleday Dell Publishing.
      Google ScholarFindings
    • Resnick, P., Iacovou, N., Suchak, M., Bergstrom, P., and Riedl, J. (1994). GroupLens: An Open Architecture for Collaborative Filtering of Netnews. In Proceedings of CSCW ’94, Chapel Hill, NC.
      Google ScholarLocate open access versionFindings
    • Resnick, P., and Varian, H. R. (1997). Recommender Systems. Special issue of Communications of the ACM. 40(3).
      Google ScholarLocate open access versionFindings
    • Reichheld, F. R., and Sasser Jr., W. (1990). Zero Defections: Quality Comes to Services. Harvard Business School Review, 1990(5): pp. 105-111.
      Google ScholarLocate open access versionFindings
    • Reichheld, F. R. (1993). Loyalty-Based Management. Harvard Business School Review, 1993(2): pp. 64-73.
      Google ScholarLocate open access versionFindings
    • Sarwar, B., M., Konstan, J. A., Borchers, A., Herlocker, J., Miller, B., and Riedl, J. (1998). Using Filtering Agents to Improve Prediction Quality in the GroupLens Research Collaborative Filtering System. In Proceedings of CSCW ’98, Seattle, WA.
      Google ScholarLocate open access versionFindings
    • Sarwar, B. M., Karypis, G., Konstan, J. A., and Riedl, J. (2000). Application of Dimensionality Reduction in Recommender System–A Case Study. In ACM WebKDD 2000 Workshop.
      Google ScholarLocate open access versionFindings
    • Sarwar, B. M., Karypis, G., Konstan, J. A., and Riedl, J. (2000). Analysis of Recommendation Algorithms for E-Commerce. In Proceedings of the ACM EC’00 Conference. Minneapolis, MN. pp. 158-167
      Google ScholarLocate open access versionFindings
    • Schafer, J. B., Konstan, J., and Riedl, J. (1999). Recommender Systems in E-Commerce. In Proceedings of ACM E-Commerce 1999 conference.
      Google ScholarLocate open access versionFindings
    • Shardanand, U., and Maes, P. (1995). Social Information Filtering: Algorithms for Automating ’Word of Mouth’. In Proceedings of CHI ’95. Denver, CO.
      Google ScholarLocate open access versionFindings
    • Terveen, L., Hill, W., Amento, B., McDonald, D., and Creter, J. (1997). PHOAKS: A System for Sharing Recommendations. Communications of the ACM, 40(3). pp. 59-62.
      Google ScholarLocate open access versionFindings
    • Ungar, L. H., and Foster, D. P. (1998) Clustering Methods for Collaborative Filtering. In Workshop on Recommender Systems at the 15th National Conference on Artificial Intelligence.
      Google ScholarLocate open access versionFindings
    Your rating :
    0

     

    Tags
    Comments