A Novel Clustering Algorithm for Detecting Community Structure in Complex Networks Based on Data Field

摘要

Network clustering for detecting community structure in networks is beneficial to understand the network structure and to analyze the network properties. In order to overcome high time complexity, difficult for the user to select initial conditions and other defects of the existing clustering algorithms, this paper analyses the above problems and proposes an adaptive clustering algorithm based on data field in complex networks. First, the importance factor is proposed to dig out the important vertices in networks as the center of clusters which is based on the merits of existing evaluation indexes of the vertex's degree, mutual information and' closeness respectively. Due to vertices in networks connected and reacted upon one another, the theory of data field in physics is introduced into complex networks, by calculating field-strength and potential function of vertices to detect clusters. Simulation experiments show that the novel algorithm can get approximate optical clusters with a lower time complexity, a higher accuracy and validity compared to other algorithms.