Opinion Dynamics Considering Social Comparison in Online Social Networks

Social comparison theory holds that people will evaluate their opinions by comparison with someone close to their own abilities or opinions. Inspired by this, we propose an opinion dynamic model referring to classic Deffuant model combined with online communication characteristics. Opinions in this model are produced along with information diffusion and thus result in dynamic population in opinion interactions. And interactions are allowed between commenters who comment on the same post or repost. Through simulations, our model is observed to present more divergent opinions than classic Deffuant model as is also believed in reference [15]. Moreover, effects of different compromise thresholds, namely confidence bound, and comparison thresholds are discussed. Further simulations are also conducted to clarify the impacts of diffusion parameters on opinion evolution.