Community discovery of complex network based on fuzzy density peak clustering algorithm

摘要

Based on the similarity measure of the complex network node and the improved density peak clustering algorithm with information entropy, a community discovery method is proposed in this paper. First of all, the distance between nodes is measured by the Jaccard similarity and the shortest path generation algorithm. Then, the core communities with the improved density peak algorithm are selected automatically. Thirdly, applying the fuzzy clustering mechanism to calculate the membership degree matrix of each point, thus completing the division of the remaining points. Finally, overlapping nodes are distinguished by setting a threshold of degree difference. Experiments are conducted on four real-life networks and the Purity and the Extended Modularity are employed to evaluate the proposed algorithm. The comparison of experiments with some classical algorithms on real networks are given to demonstrate the feasibility and effectiveness of the proposed method.