Research Of A Public Opinion Evolution Model Based On Opinion Leaders Through Game And Memory

The public opinion evolution is an important research direction of public opinion.Opinion leaders are the most influential nodes in the social opinion network,and play an important role in the evolution of public opinion.However,the classic opinion dynamics model Hegselmann-Krause,which can quickly reach the convergence of opinion views,the way to integrate opinion leaders in the improved model is a simple combination of weighted opinion values,ignoring the reasons for opinion leaders driving opinion,and no fully consider the network topology of the public opinion carrier.This leads to the difference between the evolution of the existing Hegselmann-Krause model and its improved model and the actual public opinion evolution.Based on the above phenomena,combined with the complex network,the Hegselmann-Krause model is extended,and a lyric dynamic model based on the game and memory of opinion leaders is proposed.On the improved complex network,the opinion leaders are evolved based on the principle of “Prisoner Game”.The ordinary netizens make evolution based on the HK model,and then use the memory method and node intimacy to introduce the opinion leaders' opinions into the changes of ordinary netizens' opinions.The outcome of the game determines the direction and step size of the opinion leader.During the game,the number of nodes in the bounded trust directly affects the game outcome.Aiming at the non-clustering characteristics of the classical BA scale-free network,the sub-domain and the near-field connection are introduced with different probabilities,so that the improved network has both small world characteristics and scale-free characteristics.And through simulation,get the parameters of the network characteristics similar to Twitter.This network is the carrier of the paradox dynamics model.Using Weibo's "05.30 Toutiao Tencent war" event data as the training set,using the genetic algorithm to adjust the parameters to obtain the appropriate parameters of the four models using MATLAB simulation,the results of the public opinion guidance show that: Manipulating the risk parameters of the opinions expressed by opinion leaders can control the direction of public opinion.Finally,through the microblogging modeling evolution of “10.28 Chongqing Wanzhou Gong Crashing River crash”,it is proved that the improved model better fits the actual public opinion data.