Finding Communities in Social Networks with Node Attribute and Graph Structure using Jaya Optimization Algorithm

Abstract

The complex systems can be denoted as networks where subsets of nodes are more connected among themselves. Here, the major problem is the detection of community structure in complex networks like social networks, web networks, telecommunication networks, biological, and trade networks. Community detection allows computing groups of interacting objects and the relations between them. Almost all the researches done in this area have taken into account the graph structure of the network for finding the communities. In this paper, we propose a multi objective community detection (MCD) method for complex networks using new optimization algorithm which will consider graph structure and the content associated with the nodes. Multiple metrics are used for detecting communities in these networks. These are modularity, generalized modularity and similarity of attributes. The time varying metrics are optimized by the proposed evolutionary combined Jaya optimization algorithm, which achieves exact detection without affecting execution time. The performance of proposed method has been analyzed with the five different networks. The experimental result shows that the proposed method achieves favorable results which are quite superior to existing methods.