Toward an effective hybrid collaborative filtering: A new approach based on matrix factorization and heuristic-based neighborhood

Abstract

“Collaborative filtering” (CF) methods provide a good solution for recommendation systems. Neighborhood formation is considered as the main phase in memory approaches. Unfortunately, this phase encounters many problems such as sparsity and scalability, especially for huge datasets which consists of a large number of users and items. This paper presents a new hybrid approach for collaborative filtering. It is based on two heuristic approaches for neighborhood selection. The first of which is based on selecting users who rated the same items as the active user called “intersection neighborhood”, while the second one builds the neighborhood using all users who rated one item at least as the active user called “union neighborhood”. In addition, we employ matrix factorization technique to learn the latent characteristics of the selected neighborhood (users or items) in order to quickly predict good quality of the unknown ratings. Finally, experiments show that the proposed approaches give more predictions accuracy than the traditional collaborative filtering.