A Community Detection Method Based on the Subspace Similarity of Nodes in Complex Networks

摘要: Many real-world networks have a topological structure characterized by cohesive groups of vertices. Community detection aims at identifying such and plays critical role in network science. Till now, many community methods been developed the literature. Most them require to know number communities low accuracy complex are shortcomings most these methods. To tackle issues this paper, novel method called CDNSS is proposed. The proposed based on nodes subspace similarity includes two main phases; seeding expansion. In first phase, seeds identified using potential distribution local global space. compute between each pair, specific centrality measure considering sparse linear coding self-expressiveness ability nodes. Then, with best focal state discovered which guarantees stability solutions. expansion greedy strategy used assign unlabeled nods relevant regions. results experiments performed several synthetic confirm superiority comparison well-known state-of-the-art