Mining top-k structural holes in multiplex networks

Abstract

Various real life social networks exhibit multiple types of interaction among entities, thus this arrangement builds a compilation of continuous growing networks called as multiplex networks. Lately, the area related to centrality measurement of entities in multiplex networks has developed a significant interest among researchers. However, the concept of structural holes has not received significant attention in multiplexed networks. The theory of structural holes suggests that “holes” acts as a bridge between individuals or groups that are otherwise disconnected. Structural hole plays an important role in information diffusion and link prediction. Although there are numerous methods defined to detect structural holes in a simple network, detecting structural holes in multiplexed network is still untouched. In this paper, we present a methodology of mining top-k structural holes from multiplexed or multi-layer networks. For the purpose of the study, we make use of two networks: airline and co-author. Our experiments provide an insight into the theory of structural holes in multiplexed network. We compare the proposed methodology with several alternative methods and we get encouraging and comparable results. We believe this is the first shot to report the problem of mining structural holes in multiplexed social networks.