| All along their lives, individuals take roles in their interactions with each other. This behaviour is known as the role-taking characteristic of human beings. We refer to these roles as social roles that are the primary components of societies. Identifying social roles in a society helps to better analyze the social phenomena. Consequently, it can be beneficial in the search for influentials, trustworthy people, idea innovators, etc. With this intention, we propose the structural social role mining (SSRM) framework to identify roles, study their changes, and analyze their impacts on the underlying social network. More specifically, we define four fundamental roles called leader, outermost, mediator, and outsider. Subsequently, we suggest methodologies to identify these roles within a social network. While exploring our proposed methodologies for identifying the aforementioned roles, we develop two new variants of Betweenness centrality (BC): LBetweenness (LBC) and CBetweenness (CBC). Motivated by time complexity, these two centrality measures are computed more efficiently compared to Betweenness centrality especially in large social networks. Eventually, we identify and study changes of roles in the Enron communication network using our proposed framework. According to our results, individuals serving as leaders or mediators were important people in the Enron organization. Moreover, identifying roles as well as their changes through consecutive timeframes could be informative and thus could be used as a clue for further investigations. |
