Abstract:

The division of the overlapping community structure in complex networks becomes a hot research spot, and a large number of algorithms are proposed for community structure discovery. We propose an individual conformity evolutionary algorithm (ICEA). The basic idea is to perform conformity and variation on the individual according to their probability, find the community partition with optimal (or quasi optimal) module degree in a short time, and use the neighbor voting (NV) algorithm to find overlapping nodes of the network after identifying the community structure. Experiments on real networks show that the proposed algorithm outperforms the classical algorithms in terms of time cost and division results.

Key words: complex networks, community division, evolutionary algorithm