Sequential user-item weighted-cluster extraction for Collaborative filtering

This paper proposes a new approach to collaborative filtering, in which sequential user-item cluster extraction is performed in order to relate the items to be recommended to each user. In the process, a user-item rectangular relational matrix whose elements are defined by an alternative process of “liking or not“ is first transformed into a square adjacency matrix and then co-clusters are sequentially extracted using a weighted aggregation criterion. Numerical examples including an application to a purchase history data set demonstrate the characteristics of the proposed approach.