A New Genetic Algorithm for Overlapping Community Detection

In this paper, we proposed a new genetic algorithm to detecting overlapping community in networks. This algorithm uses a new strategy to transform adjacency matrix by weighting edges according to their centrality. And we modified some rules so that genetic algorithms can be applied to detect overlapping communities. A matrix-based encoding is adopted to improve efficiency of crossover and decoding operations. We use nodes similarity to generate initial populations. By experimental comparison with other algorithms, simulation results show that the algorithm is effective and efficient when detecting overlapping community.