论文信息 - Does Wikipedia Information Help Netflix Predictions?

Does Wikipedia Information Help Netflix Predictions?

We explore several ways to estimate movie similarity from the free encyclopedia Wikipedia with the goal of improving our predictions for the Netflix Prize. Our system first uses the content and hyperlink structure of Wikipedia articles to identify similarities between movies. We then predict a user's unknown ratings by using these similarities in conjunction with the user's known ratings to initialize matrix factorization and k-Nearest Neighbours algorithms. We blend these results with existing ratings-based predictors. Finally, we discuss our empirical results, which suggest that external Wikipedia data does not significantly improve the overall prediction accuracy.

[1] Geoffrey E. Hinton,et al. Restricted Boltzmann machines for collaborative filtering , 2007, ICML '07.

[2] Yoav Shoham,et al. Fab: content-based, collaborative recommendation , 1997, CACM.

[3] Rajeev Motwani,et al. The PageRank Citation Ranking : Bringing Order to the Web , 1999, WWW 1999.

[4] Vladimir I. Levenshtein,et al. Binary codes capable of correcting deletions, insertions, and reversals , 1965 .

[5] Stephen E. Robertson,et al. Probabilistic models of indexing and searching , 1980, SIGIR '80.

[6] Daniel Lemire,et al. Slope One Predictors for Online Rating-Based Collaborative Filtering , 2007, SDM.

[7] Arkadiusz Paterek,et al. Improving regularized singular value decomposition for collaborative filtering , 2007 .

[8] Eric Horvitz,et al. Collaborative Filtering by Personality Diagnosis: A Hybrid Memory and Model-Based Approach , 2000, UAI.

[9] Yehuda Koren,et al. Improved Neighborhood-based Collaborative Filtering , 2007 .

[10] Thomas Hofmann,et al. Latent Class Models for Collaborative Filtering , 1999, IJCAI.

[11] Yehuda Koren,et al. Lessons from the Netflix prize challenge , 2007, SKDD.

[12] Luo Si,et al. Flexible Mixture Model for Collaborative Filtering , 2003, ICML.