Method Of Evolution And Reversal Of Public Opinion Based On Chinese Microblog

With the rapid development of the mobile Internet,the way people get information and participate in hot topics is changed.The emergence of mobile social platforms,such as microblog and wechat,enable people to access information,disseminate information,publish information and participate in event reviews anytime and anywhere.Microblog is the most popular and representative social platform,with hundreds of millions of users,and is also an important platform for Internet users to exchange information or ideas.In the microblog social network,users have strong influence and tendentious comments and opinions on some hot issues in real life.When users with similar attitudes continue to increase,spread and influence each other,public opinion is formed.In recent years,the frequent occurrence of network public opinion events,such as network violence,reversal events,rumors and intentional guidance events,has brought great harm to people,great negative impact to the government and mass media,as well as disharmonious and unstable factors to society.Therefore,it is more and more important to understand and master the formation and development of public opinion events,which embodies the research significance of public opinion events.A special phenomenon in the evolution of public opinion is the reversal phenomenon.Consequently,this thesis studies the evolution and reversal of microblog opinion.The main work of this paper is as follows:(1)An opinion evolution model based on three-degree influence rule is proposed.Firstly,the individual viewpoint description in the model is based on the model of continuous viewpoint and discrete behavior,which uses continuous viewpoint value to express the user's inherent tendency towards an event(internal viewpoint),and discrete behavior to express the attitude towards the event(external behavior).Among them,there is a one-to-one mapping relationship between internal views and external actions.Under the three-degree influence,the external behavior will be influenced by the two groups of opinion support and opposition.The influence of the mixing degree of two different groups of behavior directly affects the change of individuals' internal viewpoints.In this thesis,the binary information entropy theory is used to represent two different groups of behavior.The low degree of mixing indicates that one of them has great influence and is easy to change the individual's view.The high degree of mixing indicates the convergence of the two sides' influence and is not easy to change the individual's view.Secondly,through the interaction between users of microblog(including posting,forwarding and commenting),this thesis studies how the relationship network in microblog is formed.Experiments show that the formation of relationship network in microblog obeys power law distribution,so simulation experiments can be carried out on scale-free networks.Finally,the experiment analyzes the initial density of external behavior,the influence of interactive range on the evolution of viewpoints and the influence of interactive behavior on the evolution of viewpoints respectively.Case analysis show that the proposed model was reasonable.(2)This thesis proposes a microblog public opinion reversal model based on the evolution of informed individual viewpoints.Firstly,in the network,the choice of individual viewpoint is the decision-making behavior under the combined action of inner activity and external environment.Hence,continuous opinion discrete activity model(CODA model)is used to describe individual behavior.Secondly,the introduction of external forces is an important factor in the reversal of public opinion.Consequently,the use of informed individuals in the model represents the introduction of known external information.Moreover,the influence of informed individuals on the evolution model of public opinion under the same external behavior and different external behavior was analyzed in the experiment.Finally,using real data and simulation experiments to compare and explain the rationality of the proposed model in the study of public opinion reversal.This paper studies the phenomenon of reversal in the evolution and evolution of public opinion,which is to study the formation and development of public opinion events.The formation and development of research public opinion events is conducive to revealing the development process and laws of social network public opinion events,and providing theoretical support and technical support for the government to manage,control and guide the development of network public opinion in a healthy and benign direction.