Comparison of communities detection algorithms for multiplex

• Consolidate the disparate literature on the communities detection algorithms for multiplex.
• Proposed a suite of multiplex benchmark to distinguish the different algorithms.
• Empirically show the dissimilarity of the different algorithms.
• In conclusion a minor error in the assumption of a multi-relational community can significantly deviate one from their intended solution.

Multiplex is a set of graphs on the same vertex set, i.e. {G(V,E1),…,G(V,Em)}{G(V,E1),…,G(V,Em)}. It is a type of generalized graph to model the multiple relationships in a system with parallel edges between vertices. An important application in Network Science is to capture community structures in multiplex as a way to modularize the system. This paper is a literature review and comparative analysis on the existing communities detection algorithms for multiplex. The conclusion is that many of the algorithms deviate in the concept of multi-relational communities and the wrong choice of algorithm can deviate one from his intended concept.