Influence of Clustering on the Opinion Formation Dynamics in Online Social Networks

With the advent of Online Social Networks (OSNs), opinion formation dynamics continuously evolves, mainly because of the widespread use of OSNs as a platform of social interactions and our growing exposure to others' opinions instantly. When presented with neighbours' opinions in OSNs, the natural clustering ability of human agents enables them to perceive the grouping of opinions formed in the neighbourhood. A group with similar opinions exhibits stronger influence on an agent than the individual group members. Distance-based opinion formation models only consider the influence of neighbours who are within a confidence bound threshold in the opinion space. However, a bigger group formed outside this distance threshold can exhibit stronger influence than a group within the bound, especially when that group contains influential or popular agents like leaders. To the knowledge of the authors, the proposed model is the first to consider the impact of clustering capability of agent and incorporates the influence of opinion clusters (groups) formed outside the confidence bound. Simulation results show that our model can capture several characteristics of real-world opinion dynamics.