A Multi-agent PSO Algorithm to Determine the Membership percentage of Each Node in Overlapping Community in Social Networks

The social network embodies real-life social graphs. Detecting communities or clusters from these graphs is an ill-posed difficult task. The communities are identified using the adjacent nodes that have shared edges and similar features. One of the principal concerns after community detection is to identify the active nodes in a network who attend several communities. So, finding communities that are overlapped in a social network is an important topic in social network analysis. This paper introduces a multiagent algorithm particle swarm optimization. The proposed algorithm detects overlapping as well as non-overlapping communities. Following the detection of overlapping communities, this algorithm can identify those nodes leading to overlapping and ultimately it can determine the affiliation ratio of each node to the given community. The algorithm uses a special type of coding to identify the number of communities without any prior knowledge. In my method, the modularity measure is used as a fitness function to enhance particle swarm. Several tests show that the proposed algorithm which is called Fuzzy Overlapping Community Detection based on Multi-Agent Particle Swarm Optimization (FOCDMAPSO), is superior compared with four other competitive algorithms. This algorithm is implemented over six well-known datasets in the literature. Our algorithm is accomplished in detecting nodes in overlapping communities with high accurateness. Furthermore, the proposed algorithm can detect the affiliation percentage of each node that leads to overlapping communities which is a novel feature in the area of the social network.