A novel overlapping community detection using parallel CFM and sequential nash equilibrium

Detecting Overlapping Community in Social Networks is one of the challenging and complex problem. Several approaches based on heuristic, modularity & modularity density, graph partitioning and game theory are available for community detection. However getting an optimum and stable solution with less computation cost for large datasets is not possible using these existing approaches. Hence, in this work, we propose a novel overlapping community detection algorithm based on parallel community forest model and sequential Nash Equilibrium for large datasets. In this paper, community forest model (CFM) is implemented in parallel using Spark framework to get the initial community structure and then a Nash Equilibrium is computed to find a stable overlapping community structure. We conducted experiments on the benchmark LFR dataset with different sizes like 500, 1000, 2000 upto 10,000 nodes to evaluate the proposed method. Our experimental results clearly demonstrate that the proposed approach outperforms the existing works in terms of quality, scalability, stability and less computation time.