Abstract
Given a set of community detection algorithms and a graph as inputs, we propose two ensemble methods and that (respectively) identify disjoint and overlapping communities in . transforms a graph into a latent feature space by leveraging multiple base solutions and discovers disjoint community structure. groups similar base communities into a meta-community and detects both disjoint and overlapping community structures. Experiments are conducted at different scales on both synthetically generated networks as well as on several real-world networks for which the underlying ground-truth community structure is available. Our extensive experiments show that both algorithms outperform state-of-the-art non-ensemble algorithms by a significant margin. Moreover, we compare and with a recent ensemble method for disjoint community detection and show that our approaches achieve superior performance. To the best of our knowledge, is the first ensemble approach for overlapping community detection.