A Social Network Model for Analysis of Public Opinion Formation Process

The growing popularity of social networks has amplified their capacity to form public opinions. The opinion formation process is affected by social factors and social phenomena. To assess the impacts of social factors on opinion formation while considering consensus and polarity social phenomena, it is essential to employ a model that integrates the dynamics of both individuals’ connections and opinions. However, the existing social network models do not comprehensively consider the abovementioned dynamics, hence, they fail to study consensus and polarity in a single model. In this paper, we present a Directed Homophilic Preferential Attachment (DHPA) model to capture the dynamics of social network generation and rewiring (connections dynamics) and to take the variation of attitudes and characteristics of users when expressing their opinions and their desire to establish relationships with others into account (attitudes dynamics). DHPA provides the necessary facilities for examining the impact of different factors on the opinion formation process. It also enables us to analyze the circumstances for reaching consensus and polarity. It is shown that the network generated by the proposed DHPA model appropriately conforms to real social networks. We have examined the impact of some important social factors, including noisy users, censorship, and boycott campaigns, by conducting sensitivity analysis scenarios on the model, which caused interesting results. For example, why and how do noisy users prevent converging public opinion is shown. Also, we found that censorship on social networks can significantly undermine the formation of consensus in public opinion.