Detecting overlapping communities in online social networks using game theoretic approach

In today's world, online social networks (OSNs) such as Twitter and Facebook capture interactions among people through comments on blogs, posts and feeds. E-marketing, popularized by the development of OSNs, relies heavily on the structure of the social networks where customers are represented by nodes and their interactions as edges. Social networks contain community structure which means people have much denser connections within the groups than between the groups. Since people have more than one characteristic, detecting overlapping communities becomes an important issue. Our focus in this paper is on identifying significant attributes which have a strong relationship to the property of communities such as size and density strongly supported by statistical methods. We discuss game theoretic community detection algorithm and validate it by running experiments on Facebook data and extract dominant attributes in the communities.