Multiplex community detection in complex networks using an evolutionary approach

Abstract

Multiplex networks are the general representative of complex systems composed of distinct interactions between the same entities on multiple layers. Community detection in the multiplex networks is the problem of finding a shared structure under all layers, which combines the information of the entire network. Most of the existing methods for community detection in the single-layer networks cannot be well applied to detect shared communities in multiplex networks. In this paper, we employ a multi-objective evolutionary approach, namely Multi-Objective Evolutionary Algorithm based on Decomposition with Tabu Search (MOEA/D-TS), to detect shared communities in multiplex networks. Also, we have improved the MOEA/D-TS using a social networks analysis measure named Clustering Coefficient (CC) in terms of the generation of the initial population. This hybrid algorithm employs the parallel computing capacity of the Multi-Objective Evolutionary Algorithm based on Decomposition (MOEA/D) along with the neighborhood search authority of Tabu Search (TS) for discovering Pareto optimal solutions. Extensive experiments on a variety of single-layer and multiplex real-world data sets show the superiority of the proposed method in comparison to state-of-the-art algorithms and its capability for producing improved results.