Spanning Tree Based Community Detection using Min-Max Modularity

Community refers to the group of entities which have similar behavior or characteristic among them. Usually community represents basic functional unit of social network. By understanding the behavior of elements in a community, one can predict the overall feature of large scale social network. Social networks are generally represented in the form of graph structure, where the nodes in it represent the social entities and the edges correspond to the relationships between them. Detecting different communities in large scale network is a challenging task due to huge data size associated with such network. Community detection is one of the emerging research area in social network analysis. In this paper, a spanning tree based algorithm has been proposed for community detection which provides better performance with respect to both time and accuracy. Modularity is the well known metric used to measure the quality of community partition in most of the community detection algorithms. In this paper, an extensive version of modularity has been used for quality assessment. (C) 2016 The Authors. Published by Elsevier B.V.