Augmenting Collaborative Recommenders by Fusing Social Relationships: Membership and Friendship

Collaborative filtering (CF) based recommender systems often suffer from the sparsity problem, particularly for new and inactive users when they use the system. The emerging trend of social networking sites can potentially help alleviate the sparsity problem with their provided social relationship data, by which users’ similar interests might be inferred even with few of their behavioral data with items (e.g., ratings). Previous works mainly focus on the friendship and trust relation in this respect. However, in this paper, we have in-depth explored a new kind of social relationship - the membership and its combinational effect with friendship. The social relationships are fused into the CF recommender via a graph-based framework on sparse and dense datasets as obtained from Last.fm. Our experiments have not only revealed the significant effects of the two relationships, especially the membership, in augmenting recommendation accuracy in the sparse data condition, but also identified the outperforming ability of the graph modeling in terms of realizing the optimal fusion mechanism.