Abstract:

Overlapping community structure detection bears both theoretical and practical significance for the study of complex systems. We propose an overlapping community detection algorithm based on local expansion optimization. Firstly, a group of irrelevant seeds with large clustering coefficient are selected as initial communities according to the clustering coefficient of network nodes. Then, the initial communities are expanded into tightlyconnected local communities by a greedy strategy. Finally, similar communities are merged and overlapping community structures with high cover rate are obtained. Experimental results on both synthetic and realworld networks show that compared with other representative local expansion methods, the proposed algorithm can efficiently detect overlapping communities of higher quality in the networks with different sparsity degrees.

Key words: complex network, overlapping community detection, local expansion, structural fitness