A new local algorithm for overlapping community detection based on clustering coefficient and common neighbor similarity.

Identification of communities represents one of the most important fields in social network analysis. In this paper, we propose a new local algorithm mainly focused on finding the initial communities then expanding them by using a new weighted belonging degree. The communities are selected according to the local clustering coefficient and the common neighbors similarity of their members, which develop the level of uncovering overlapping communities. The overlapping modularity and the F-score are exploited to evaluate the quality of the results. The experiments show significant improvement while using the new weighted belonging degree.