Evolution of communities in dynamic social networks: An efficient map-based approach

Abstract

The expanded domain of expert system applications has risen the impact of modeling and analysis of community evolution in social networks as an important part of the decision-making process. Social networks are time-variant systems, evolving through entities joining or leaving networks and establishing or terminating relationships. In this article, we study evolution of social networks at the level of community structure, by tracking different transformations of communities over time. Upon experimentation, we observed that a considerable portion of community evolution is partial events such as partial merge. Therefore, we define a broader set of community evolution to include partial events. Furthermore, we introduce ICEM, a novel method for Identification of Community Evolution by Mapping. ICEM determines community evolution by tracking community members, implemented with a hash-map. ICEM maps each member to a (t, c) pair, specifying it is last observed in time window t and community c. We evaluated our proposed approach with seventeen publicly available social network datasets and compared its performance against other well-known methods in the literature. Our experimental results indicated the performance superiority of our proposed solution. Additionally, we conducted separate comprehensive experiments using three community detection algorithms to highlight the effect of choosing different community discovery methods on community evolution results.