Collaborative Filtering by User-Item Clustering Based on Structural Balancing Approach

Collaborative filtering is a technique for reducing information overload and is achieved by predicting the applicability of items to users. In neighborhood-based algorithms, the applicability is predicted by the weighted averages of ratings of neighbors. This paper considers a new approach to user-item clustering in collaborative filtering. The new clustering method plays a role for selecting the user-item neighbors based on a structural balance theory used in social science, in which users and items are partitioned into two clusters by balancing a general signed graph composed of alternative evaluations on items by users.