A variance-aware multiobjective Louvain-like method for community detection in multiplex networks

Abstract In this article, we focus on the community detection problem in multiplex networks, that is, networks with multiple layers having same node sets and no inter-layer connections. particular, look for groups of nodes can be recognized as communities consistently across layers. To end, propose a new approach generalizes Louvain method by (a) simultaneously updating average variance modularity scores (b) reformulating greedy search procedure terms filter-based multiobjective optimization scheme. Unlike many previous maximization strategies, which rely some form aggregation various layers, our aims at maximizing individual modularities each layer simultaneously. We report experiments synthetic real-world showing effectiveness robustness proposed strategies both informative case, where all show structure, noisy represent only noise.