Detection algorithm based on closeness rank and signal transimission

Abstract

In the algorithm of community detection using clustering technology, the prior information of community structure and the similarity measure between nodes influence the clustering effect greatly. For how to adaptively discover the community and utilize the network topology to calculate the similarity between nodes in order to raise the modularity of network, the community detection algorithm based on closeness ranking and signaling transmission has been proposed in this paper. The main idea of the algorithm is to rank the closeness centrality of nodes, select the initial center nodes according to the certain rules. The similarity between nodes are calculated according to the idea of signal transmission, and each node in the community is assigned to the candidate community. Finally, Merging of small community are carried out adaptively, and the center nodes set are updated iteratively until the community is stable. Experiments are carried out on Zachary Karate Club Network, American College Football, Dolphin Social Network, Lesmis vocabulary network, and the results demonstrate the feasibility and efficiency of the proposed algorithm.