An Evolutionary Algorithm for Community Detection Using an Improved Mutation Operator

Community detection aims to determine topological structures and detect patterns in complex networks, presenting a fundamental and important problem in computer science. Community detection problem can be defined as an NP-hard combinatorial optimization problem, where recently several optimization models have been developed in the literature. Most studies adopted the evolutionary algorithm to resolve the community detection problem by optimizing a single objective function along with traditional evolutionary operators. The predictive power of the evolutionary-based community detection algorithm can be improved by incorporating domain-specific knowledge in evolutionary operators to make the exploration process of the search space more effectively and thus produce better solutions. To this end, this paper aims to present an evolutionary algorithm addressing community detection problem more reliable by applying an improved mutation operator based on the definition of the weak community along with optimizing community score model as a fitness index. Results of systematic experiments on three real complex networks have demonstrated the effectiveness of the proposed algorithm.