Community discovery based on social actors' interests and social relationships

The increasing availability of social network data has motivated more computational research in social network analysis. Recently, discovering community from social networks came into the big picture of social network study. In this paper, we propose a novel approach for community discovery based on the contents of social actors' personal interests and their social relationships. Our dataset is a populated ontology created by integrating the publicly available FOAF (Friend-of-a-Friend) documents. We firstly discover the communities based on the social actors' personal interests. Then we use the social relationship to extend the communities for further discovery. Finally we conduct extensive experimental study to demonstrate that our approach is effective for community discovery.