Research On The Evolution,Prediction And Response Strategy Of Public Opinion Of Accident And Disaster Network

The popularity and deep development of the Internet has changed the way people communicate information.With the seamless embedding of communication technology,the network virtual world and the real world can interact in real time,and the social activities of human beings can be reflected in the network in time.The influx of all kinds of network platforms provides netizens with a place for news reports,exchange of views and emotional expression.At the same time,when the accident disaster events caused by man-made causes come to human society,under the condition of high network penetration,accident disaster network public opinion has seriously intervened in peopleundefineds lives,changed the network ecological environment,and influenced the progress of real events.It will bring certain risks to the network ecology,and it is urgent for public opinion to deal with it scientifically.Therefore,how to grasp the evolution character and the transmission law of the accident disaster network public opinion and realize the accurate prediction of the public opinion evolution to provide the theory and the method foundation for the government scientific response,has become an important public security problem that the government of our country faces,and has become the academic circle,The focus of the public and the government.In view of this,this paper,based on the theory of life cycle and the theory of government emergency management,adopts a variety of quantitative methods such as social network analysis,stochastic forest algorithm and so on.The general characteristics of public opinion evolution of accident disaster network,the characteristics of public opinion evolution in each stage,the prediction of public opinion evolution and the government coping strategies are studied deeply.First of all,this paper makes an in-depth analysis of the evolution process of accident disaster network public opinion from the micro and macro point of view.in order to analyze the evolution characteristics,propagation law and internal mechanism of public opinion evolution.This paper puts forward the evolution idea of public opinion based on life cycle,divides its evolution process into three stages,analyzes the theme and emotional attribute of public opinion in each stage,and completes the overall evolution analysis of public opinion from the macro aspect.the social network analysis method is applied to the public opinion transmission process,the role of the public opinion participants in the exchange of the public opinion information and the information relation between them is analyzed,the structural characteristics of the social network and the central character of the user node are analyzed for the public opinion information,The analysis of the evolution of public opinion is completed from the micro aspect.Then,based on the overall data volume of the public opinion,the stochastic forest algorithm(RF)was introduced,and the prediction of the public opinion evolution was completed from three aspects.firstly,establishing a prediction model,building an RF prediction model through a CART regression tree and a bagging integration algorithm;secondly,carrying out experiment and simulation,carrying out training on the acquired data by using an R language platform,and forming a better model after a plurality of times of debugging;Third,the performance of the model is compared.the common multiple linear regression and BP neural network prediction methods are compared with RF model,and the difference between their predicted value and real value is analyzed,and several error values are used as reference basis to highlight the advantages of RF prediction model.Finally,based on the potential risk of public opinion evolution,the government control model is set up,and the government response strategy of public opinion is put forward.Excavating the potential risk of accident disaster network public opinion,regulating and controlling the network public opinion of accident disaster from three aspects: consciousness training,technical support and guarantee system,and then combining with the life cycle theory,puts forward the public opinion response scheme of different evolution stages,which provides a reference for the government to deal with the accident disaster network public opinion scientifically.In this paper,life cycle theory,social network theory and random forest algorithm are introduced into the construction of accident disaster network public opinion response,and the theory and decision-making method of systematic public opinion evolution analysis,prediction and government response are established in order to solve the weak problems of government scientific response in the current accident disaster network public opinion research,which is expected to be in the development of network public opinion theory system and the improvement of government accident disaster public opinion response ability.The development and construction of intelligent response system of accident disaster public opinion and the improvement of efficiency and benefit of emergency management of accident disaster network public opinion play an important role in promoting the development and construction of intelligent response system of accident disaster public opinion.