Enhanced overlapping community detection in social networks using wise initialization

Social networks represent relationships between people. Communities are groups of people with some common interests or features. In this context, overlapping means a community member could be member of some other communities at the same time. In this paper, we introduce a new framework to enhance the performance of overlapping community detection techniques. In this method, the target network is divided into several subnets and after detecting their communities, this information is used as an initialization for the final community detection technique.