An Improved Grey Wolves Optimization Algorithm for Dynamic Community Detection and Data Clustering

One of the salient features of real-world networks such as social networks is the existence of community structures. Because of the importance of groups and communities in social networks, various algorithms have been proposed to identify communities in this type of dynamic networks. In this paper, we present a new approach to community recognition in dynamic social networks, which is multi-objective and metaheuristic. Our approach is to improve the Grey Wolf Optimizer algorithm and the Label Propagation algorithm and to combine the two algorithms for better performance. We performed our experiments on two artificial and real datasets, and the results show that our proposed method performs better compared to present algorithms in terms of both quality and detection speed. We also applied our proposed algorithm to 23 base functions, which performed better than the other metaheuristic algorithms. At the end, the performance of our proposed algorithm is compared to six other clustering methods on nine datasets from the UCI machine learning laboratory. The simulation results show the effectiveness of the proposed algorithm for solving data clustering problems.