Study On The Method Of Public Opinion Guidance Based On Microblog

With the rapid development of Internet technology and the penetration of free views into the hearts of the people,more and more social platforms have become a new channel for public opinion release and dissemination.Microblog,with its good interaction and convenient news dissemination,has become the main platform for public opinion dissemination.In recent years,with many emergencies and hot topics are first revealed through Microblog platform,more and more citizens also consciously express their demands through Microblog,sticker bar and other forms to elaborate their feelings.Under the background of big data of Microblog platform,faced with massive data,how to discover the law of public opinion dissemination,identify the public opinion accurately,and guide the negative public opinion to the positive public opinion independently has become the hot spot of public opinion research.However,the research on the guidance of public opinion is mainly focused on communication,journalism and other subjects.The way of using computer guidance is relatively single,the guiding idea lags behind,the efficiency is low,and the form of supervision is lack of automation.In view of this phenomenon,this paper studies the factors,rules and guiding methods of public opinion dissemination in Microblog.The concrete work is as follows:1.According to the theory of herd behavior and other theories,this paper defines and quantifies the factors affecting public opinion communication.Firstly,the influence of user nodes is calculated by the forwarding,commenting and praising in the process of public opinion dissemination.Then,we quantify the topic popularity with the improved Newton's law of temperature cooling and the number of interactions generated in the topic,analyze the relationship between the topic popularity and time.Finally,the topic interest is defined and quantified according to the similarity between the user's historical posts and the current public opinion topic.2.In this paper,a public opinion evolution model HK-SEIR,which combines the view fusion model HK and the epidemic transmission model SEIR,is proposed to analyze the rule of public opinion evolution.Firstly,the interest degree threshold is added to the view fusion model HK,and the topic interest and user influence are introduced in the process of view fusion,and the interaction behavior between the users under the interest and confidence threshold is analyzed.Then,we calculate the probability of topic propagation caused by the interaction of views between users under group pressure,and the probability that users change from infected state to removed state under topic popularity.Then,the HK-SEIR model is constructed by integrating the view fusion model and the epidemic model on the calculated probability,and analyzed the changes of topic susceptible,exposed,infected and removed in the process of public opinion communication3.Analyze the network characteristics of each life cycle of public opinion,and construct a public opinion guidance method using HK-SEIR model,structural hole and homogeneity.Firstly,the network characteristics of the life cycle of public opinion transmission are analyzed through the density of infected I and removed R of HK-SEIR model,structure hole,network homogeneity and the structure of public opinion network.Then,the neural network is used to quantify the importance of the structural hole nodes by the characteristic parameters such as the constraint coefficient(CT),data centrality(BT)of the effective scale(ES),etc.Finally,the network structure of the life cycle of public opinion is constructed according to the infected node density and removed node of HK-SEIR model,structural holes and homogeneity of public opinion network.and formulate the guiding strategy.By adjusting the parameters of structural and non-structural hole nodes,the growth and maturity of public opinion in the process of dissemination are accelerated to change into recession.Finally,in order to verify the validity of the proposed models and algorithms,the experiment is simulated and analyzed in the real Weibo dataset.The experiment proves that the HK-SEIR model is closer to the real network communication law of public opinion than the SEIR,SIR model.It is also more consistent with the changing trend of the topic of public opinion dissemination and the rule of work and rest of the individual communication of public opinion.It is also found that interest is the main factor influencing the communication of public opinion after the interaction of individual views reaches a certain degree.The higher the negative proportion of public opinion,the higher the peak value of public opinion,and the longer the public opinion lasts.In the public opinion guidance experiment,it is found that changing the viewpoint of negative communication homogenous group node can reduce the peak value of public opinion transmission and promote the public opinion to enter the recession period ahead of time.When the ratio of structural tunnel nodes ? [0.2,0.3],and the proportion of unstructured tunnel nodes ? [0.2,0.4] are changed,the guiding effect is better.And in the growing period of public opinion,the effect of public opinion guidance is better than the mature period of public opinion guidance.