Influence Ranking Model for Social Networks Users

Abstract

Microblogging is the most important feature for Social Networks (SN) nowadays. It allows users to interact together by sharing and posting contents. The concept of spreading content between users raises an important question: Who are the users responsible for this content? In other words, the detection of content spreaders becomes one of the most important analytic issues. The common belief is that the best content spreaders are the best connected users (the most central users within network). Specifically, k-shell decomposition methodology defines the most efficient content spreaders as those located within the core of the network. In this paper, influence ranking model (IRM) is presented to rank SN users based on their contribution in spreading a specific content. The proposed model is inspired by the pruning process of the powerful k-shell decomposition methodology. IRM has been evaluated in realistic experiments using the famous datasets of Advogato trust network and Bitcoin Alpha trust weighted signed network. The proposed model was assessed in terms of distinction of nodes ranking and dissemination capability. Results have shown that IRM has promising results in SN users ranking.