A local information based multi-objective evolutionary algorithm for community detection in complex networks

Abstract

Due to the important role in analyzing the structure and function of complex networks, community detection has attracted increasing attention in the past years. Multi-objective evolutionary algorithms (MOEAs) have shown promising performance in community detection and, in this paper, we continue this research line by further exploring the potential of MOEAs in detecting communities. To be specific, a local information based MOEA, termed LMOEA, is proposed for community detection, where an individual updating strategy is suggested to improve the quality of community detection. Considering that a network often contains some local communities which are easily detected in the early evolutions, the proposed strategy utilizes these local communities found by individuals to guide the search in the following generations. The effectiveness of the proposed LMOEA is verified by comparing it with several existing evolutionary algorithms for community detection on both synthetic and real-world networks. Experimental results demonstrate the competitiveness of the proposed LMOEA for community detection in complex networks.