An ideal point based many-objective optimization for community detection of complex networks

•A novel many-objective community detection method called MaOCD is proposed.•An ideal point based strategy is used to improve separability of solutions while many objectives are optimized.•MaOCD uses the locus-based representation to represent and initialize solutions.•A novel metric is introduced for comparing multi/many-objective community detection algorithms.•The results show that MaOCD performs better than single/multi-objective community detection methods.