Multi-level Module Identification Method for Directed Complex Network

Abstract

Module identification of directed complex networks is an important research content in the field of complex networks. In the past, complex network overlapping module recognition algorithms based on node feature attributes could not effectively identify overlapping modules in a directed network. This paper designs a multi-level directed complex network overlapping module identification algorithm. The algorithm is divided into two levels. The low-level algorithm calculates the similarity of edges and edges based on the edge feature, and the high-level algorithm calculates the feature values of the edge based on the distribution characteristics of the triples. Then we convert the directed node network into an undirected weight edge network. Next, the edge network is clustered to identify the overlapping module structure. In this paper, the algorithm is applied to the transcriptional regulation networks. The experimental results show that the algorithm can accurately identify the overlapping module structure in the directed complex network.