A Genetic Algorithm Hybridized with an Efficient Mutation Operator for Identifying Hidden Communities of Complex Networks

Community detection is main research trend in analyzing of complex networks with aiming at uncovering their structural properties. Recently, many evolutionary methods were employed to identify communities of complex networks and achieved efficient results. The main challenge regarding the application of evolutionary-based approaches, specifically to handle large complex networks, is relatively long execution time and lower convergence speed. In this paper a novel evolutionary method called EGACD for uncovering hidden communities of complex networks is proposed. Our method hybridized a novel local search strategy along with the locus-based representation to accelerate the convergence and improve the accuracy. This type of representation reduced the search space and incorporates domain-specific knowledge with the solutions through initialization and reproduction operators. The proposed method does not requires knowing the number of communities at the beginning of the search process. The experiments with the real-world and LFR network datasets demonstrate the relatively high capacity of our proposed genetic algorithm in detecting communities with relatively lower generations and more precision.