A Bidirectional Collaborative Filtering Recommender System Based on EM Algorithm

Collaborative filtering is a promising recommendation technique for predicting the preferences of users in recommender systems. The date coming from recommender system is not only big but also sparse. It motivates the need for a more intelligent approach to obtain the tastes of users. In this paper we present a novel bidirectional collaborative filtering method based on EM algorithm. We combine user rating with item rating to build a new rating matrix and calculate a steady result by iterating this strategy. The empirical evaluation demonstrates that our technique have encouraging result contrast to the traditional approaches.