Where Does My Brand End? An Overlapping Community Approach

Abstract

In this paper, we present a new Memetic Algorithm for overlapping community detection. We use a link-based clustering approach to detect the communities of edges in complex networks. To assess the quality of our method, we present experimental results for benchmark networks in comparison to other state-of-the-art algorithms. In addition, we present a case study of a co-purchasing product network from a brand-centric point of view to show the real-life utility of this new Memetic Algorithm.