Community Detection in Attributed Graphs with Differential Evolution

Detecting communities in networks, by taking into account not only node connectivity but also the features characterizing nodes, is becoming a research activity with increasing interest because of the information nowadays available for many real-world networks of attributes associated with nodes. In this paper, we investigate the capability of differential evolution to discover groups of nodes which are both densely connected and share similar features. Experiments on two real-world networks with attributes for which the ground-truth division is known show that differential evolution is an effective approach to uncover communities.