Matching Users with Groups in Social Networks

Understanding structures and dynamics of social groups is a crucial issue for Social Network analysis. In the past, several studies about the relationships existing between users and groups in On-line Social Networks have been proposed. However, if the literature well covers the issue of computing individual recommendations, at the best of our knowledge any approach has been proposed that considers the evolution of on-line groups as a problem of matching the individual users’ profiles with the profiles of the groups. In this paper we propose an algorithm that addresses this issue, exploiting a multi-agent system to suitably distribute the computation on all the user devices. Some preliminary results obtained on simulated On-line Social Networks data show both a good effectiveness and a promising efficiency of the approach.