Overlapping community detection through an improved multi-objective quantum-behaved particle swarm optimization

摘要: Community detection is one of the most important problems in field complex networks recent years. The majority present algorithms only find disjoint communities, however, community often overlap to some extent many real-world networks. In this paper, an improved multi-objective quantum-behaved particle swarm optimization (IMOQPSO) based on spectral-clustering proposed detect overlapping structure Firstly, line graph modeling network formed, and a spectral method employed extract information graph. Secondly, IMOQPSO solve problem so as resolve separated which corresponding presenting network. Finally, fine-tuning strategy adopted improve accuracy detection. experiments both synthetic demonstrate our achieves cover results fit real situation even better fashion.