Agent-based simulation research on group emotion evolution of public emergency

Abstract

The use of agent based social simulation in computational analysis of public emergency management is becoming a prominent approach in social science research. Due to the complexity of rumor spreading and the radical emotion infecting processes, the results of the group emotion evolution analysis often lose effectiveness, which will affect the control and strategy establishment in emergency management. In this paper, we study the group emotion evolution process with the agent based simulation method. Along the way, the events-chain model is proposed to modeling the over-lapping influence of the temporal successive events. The simulation experiments of correlation events and events-chain model of a real public incident are performed based on the constructed framework. The results show that the group emotion evolution processes of correlation events have different patterns with the single event, and they cannot analyze as two independent events. Meanwhile, the results of real incident show that the events-chain model can simulate the real group response with consistency of the evolution.