Research on Motif Mining Based on Neighborhood Equivalence Class

Abstract

Abstract The motif is an important mesoscopic structure existing in the network, and motif mining is an important means to study the social network structure. Based on the tree traversal G-tries algorithm of common subgraphs, we propose an accurate subgraph recognition algorithm of neighborhood equivalence class Ex-Motifs to reduce the matching process of subgraph isomorphism. In addition, for the research of motif metric, we propose a motif metric index based on a common substructure, which can directly judge the significance of subgraph frequency on the original network. Experimental results show that the computational efficiency of Ex-Motifs is relatively high, and it can find a motif similar to the traditional motif metric method.