Extending opinion dynamics model for collective online behaviours analysis

Shixiong Wang, Yi Jiang

Department of Management Science and Engineering, Zhejiang Sci-Tech University, 310018, Hangzhou, China

In online social networks, opinion dynamics generally lead to different types of collective online behaviour such as consensus, polarization and fragment. Then an open problem arises: how are different typical collective online behaviours emerged from the behavioural decisions of individual and interactions among individuals during the process of opinion dynamics? This work examines the process of opinion dynamic in online social networks and different types of interactions among individuals on this process. An opinion-driven dynamics model, which combines a social network-based opinion dynamics model with generative individual behaviour, is proposed by adding antagonistic responses to the DW model. The proposed model integrates three types of interactions and setting up two thresholds to characterize individual behaviour. The behavioural component utilizes an initiation threshold such that if an individual's opinion exceeds this threshold, the individual will initiate the behaviour. In order to verify the effectiveness of the model, simulations are presented to examine how different typical collective behaviours emerge. As a result, we find that opinion dynamics with different threshold lead to different types of collective online behaviours. The openness of individuals to a differing opinion is the key factors to consensus or fragment.