Clustering for probabilistic model estimation for CF

ABSTRACT

Based on the type of collaborative objects, a collaborative filtering (CF) system falls into one of two categories: item-based CF and user-based CF. Clustering is the basic idea in both cases, where users or items are classified into user groups where users share similar preference or item groups where items have similar attributes or characteristics. Observing the fact that in user-based CF each user community is characterized by a Gaussian distribution on the ratings for each item and the fact that in item-based CF the ratings of each user in item community satisfy a Gaussian distribution, we propose a method of probabilistic model estimation for CF, where objects (user or items) are classified into groups based on the content information and ratings at the same time and predictions are made considering the Gaussian distribution of ratings. Experiments on a real-world data set illustrate that our approach is favorable.

REFERENCES

• 1: Badrul Sarwar , George Karypis , Joseph Konstan , John Reidl, Item-based collaborative filtering recommendation algorithms, Proceedings of the 10th international conference on World Wide Web, p.285-295, May 01-05, 2001, Hong Kong, Hong Kong [doi>10.1145/371920.372071]
• 2: Qing Li , Byeong Man Kim , Dong Hai Guan , Duk whan Oh, A music recommender based on audio features, Proceedings of the 27th annual international ACM SIGIR conference on Research and development in information retrieval, July 25-29, 2004, Sheffield, United Kingdom [doi>10.1145/1008992.1009106]
• 3: Qing Li , Byeong Man Kim, Clustering Approach for Hybrid Recommender System, Proceedings of the IEEE/WIC International Conference on Web Intelligence, p.33, October 13-17, 2003
• 4: Breese, J. S., Heckerman, D. and Kardie, C. (1998). Empirical Analysis of Predictive Algorithms for Collaborative Filtering. In Proc. Of the 14th UAI, pp.43--52.
• 5: Olfa Nasraoui , Mrudula Pavuluri, Accurate web recommendations based on profile-specific url-predictor neural networks, Proceedings of the 13th international World Wide Web conference on Alternate track papers & posters, May 19-21, 2004, New York, NY, USA [doi>10.1145/1013367.1013445]

INDEX TERMS

Primary Classification:
  H. Information Systems
  H.3 INFORMATION STORAGE AND RETRIEVAL
      H.3.3 Information Search and Retrieval
          Subjects: Information filtering

General Terms:
Algorithms, Experimentation

Keywords:
collaborative filtering, information filtering, probabilistic model