SSCF: A semantic and social-based collaborative filtering approach

The present paper concerns the members recommendation in social networks. The proposed approach is based on the semantic and social collaborative filtering technique (SSCF). In this approach, the formation of communities of users is based on the calculation of similarities between them and includes semantic and social dimensions. These two dimensions are respectively related to the calculation of similarity between the user and (1) his close friends and (2) those he trusts. A recommender system based on this approach has been developed. The preliminary experiments results show the importance of integrating the semantic and social dimensions in the recommendation process.