WOCDA: A whale optimization based community detection algorithm

Community of complex networks is one of the most important properties in networks, in which a node shares its most connections with other nodes in the same community. Community detection, which can be viewed as an optimization problem, has received a lot of attention in the field of complex networks. Whale Optimization Algorithm (WOA), a recently proposed meta-heuristic algorithm, is designed to mimic the hunting behavior of humpback whales and deal with the optimization problem. In this paper, a new community detection algorithm, Whale Optimization based Community Detection Algorithm (WOCDA), is proposed to discover communities in networks. In WOCDA, a new initialization strategy and three operations, shrinking encircling, spiral updating and random searching, are designed to mimic the hunting behavior of humpback whales. Firstly, the initialization strategy with label diffusion and label propagation is developed to obtain the high-quality initial solution. Then shrinking encircling operation based on label propagation is built to update the label of the current node with the label of its most neighboring nodes own. After that, spiral updating operation is established to keep good communities by the one-way crossover operator. Finally, random searching operation is created to randomly choose the label of a neighboring node and update the label of the current node so as to increase the ability of global search. Experimental results on synthetic and real-world networks demonstrate that WOCDA can successfully detect communities and obtain more accurate results than state-of-the-art approaches. (C) 2019 Elsevier B.V. All rights reserved.