Social Network Analysis to Detect Inherent Communities Based on Constraints

Abstract

Social network analysis (SNA) is used to analyze social networks or structures made up individuals called nodes, which are tied by one or more specific types of interdependency such as relationships, connections, or interactions. Often it is used in many internet-based applications like, social networking websites, on-line viral marketing, and recommendation network based applications to improve the performance of user-specific information dissemination. Detecting communities, which are basically sub-graphs or clusters, within a social network has been the central focus of this work. Here, we present a divisive hierarchical clustering algorithm for detecting disjoint communities by removing minimum number of edges to obey minimum edge-cut principle, like CHAMELEON: Two Phase Agglomerative Hierarchical Clustering. The stopping criteria of this algorithm depends on two threshold constraints namely, balance constraint (BC) and MINSIZE (MS) like CHAMELEON. As a measure of the quality of community, we follow network centrality measure clustering coefficient. Our experimental results, using some well-known benchmark social networks, also show that our method determines similar communities with good average clustering coefficient as the other existing well known methods of various research papers.