A comprehensive view of community detection approaches in multilayer social networks

Multilayer social networks are the main representative form for today’s social networks. In fact, the multiplicity of relations, the huge amount of data and the dynamic nature of nowadays social networks impose the representation of the network with multiple layers. This new representation makes network analysis more challenging especially Community retrieval. So, researchers propose different approaches to handle these challenges to detect accurate communities in the multilayer networks. The main goal of this paper is to present a novel and comprehensive view on community detection strategies within multilayer social networks. To do so, we provide a taxonomy of existing methods in static and dynamic multilayer social networks. Additionally, we introduce a "four worlds framework" to offer a comprehensive comparison of the different existing community detection methods. Lastly, we outline potential avenues for future research and highlight some unresolved challenges in this domain.