Abstract This study makes attempts to forecast the trend of public opinion and analyses the different modes of governance that government deals with network mass incidents using evolutionary game theory. The diffusion function of public opinion in network mass incidents is built by using the generalized replicator dynamics model, where the strategy interaction and behavior evolution between local government and social vulnerable groups are put into the traditional infectious disease diffusion model to simulate the spread and convergence process of public opinion. It is concluded that enlarging the probability of obtaining government compensation will cause the mass incidents, and meanwhile, to slow down the spread of public opinion government needs to control the opening communication degree and the protester number in the explosion phase. And in order to make the public opinion calm and maximize the social welfare, the high-level government needs to set the punishment mechanism and keep communication open.
2015, Vol. 12 
