Learning Collaborative Information Filters

TLDR

Predicting user preferences from other users’ ratings is a common recommendation strategy, yet existing algorithms do not leverage machine learning literature. The authors aim to create a representation for collaborative filtering that enables the use of any machine learning algorithm. They propose using singular value decomposition on the initial rating matrix to capture latent structure, eliminating the need for users to rate common items. On a large movie rating database, the method significantly outperforms current collaborative filtering algorithms.

Abstract

Predicting items a user would like on the basis of other users’ ratings for these items has become a well-established strategy adopted by many recommendation services on the Internet. Although this can be seen as a classification problem, algorithms proposed thus far do not draw on results from the machine learning literature. We propose a representation for collaborative filtering tasks that allows the application of virtually any machine learning algorithm. We identify the shortcomings of current collaborative filtering techniques and propose the use of learning algorithms paired with feature extraction techniques that specifically address the limitations of previous approaches. Our best-performing algorithm is based on the singular value decomposition of an initial matrix of user ratings, exploiting latent structure that essentially eliminates the need for users to rate common items in order to become predictors for one another's preferences. We evaluate the proposed algorithm on a large database of user ratings for motion pictures and find that our approach significantly outperforms current collaborative filtering algorithms.