Effective Community Detection Algorithm Based on Edge Influence Weight

Abstract

Connections strength between nodes are fundamental and important components in social networks, and connection strength determines the community structure of the network to a large extent. Edge weight is a meaningful representative of connection strength or data credibility, which can be applied to social network analysis. Aiming at the problems of insufficient research on the relationship between nodes and unreasonable initial selection of community centers, a community detection algorithm based on edge influence weight (CDP-EW) was proposed in this research. Specifically, to solve the initial community center selection problem, the degree centrality of nodes was used to calculate node influence. Then, the edge influence weight was redefined to calculate similarity based on the link relationships between nodes. Moreover, CDP-EW was compared with some community detection algorithms on real complex network datasets in experiments, where the proposed algorithm performed well on complex networks.