Journals Proceedings

Abstract

Community detection is a vital research area for online social networks. Since there is not a formal context in users` profiles, a new data source of user`s attributes is extracted from online social networks. Then in this paper a novel algorithm is investigated. Both attributes and relations of users are used in the proposed algorithm; therefor e communities can be detected through users’ similar characteristics or users common relationships. The experiments show that the accuracy of the algorithm is comparable to other well-known algorithms; moreover detected communities are self-descripted through the mode of each community members