Data-driven Research And Application On The Modeling Of Network Public Opinion Propagation And Parameter Inversion

With the vigorous development of the Web2.0,the network has gradually become the most active expression of public opinion,and the network public opinion involving various social phenomena and social events could evolve into an issue in politics and public administration.In this context,it is of great practical significance to study the evolution law and the propagation process of the actual social network public opinion.A network opinion propagation model is constructed by improving the classical disease spreading SIR model.Concern these problems that the existing researches on public opinion propagation model are seldom combined with the actual data and digging out the inherent law of public opinion propagation from the opinion big data becomes an urgent problem,a parameter inversion algorithm of the public opinion propagation model using neural network is proposed based on the practical social opinion big data.Based on Sina micro-blog big data analysis and the proposed model,the parameter inversion algorithm can be used to predict the network public opinion's trend of actual cases.Then compare the parameter inversion algorithm with the Markov prediction model.Firstly,the current situation of the network public opinion propagation is discussed.On the basis of introducing the complex network theory,the network public opinion and the big data,the deficiencies of the existing research results in the network public opinion propagation are pointed out.Secondly,on the basis of analyzing current infectious disease model and opinion dynamics model,a network opinion propagation model based on actual data is constructed by improving the classical disease spreading SIR model combined with the characteristics of the actual social network public opinion.Thirdly,in order to dig out the inherent law of public opinion propagation from the opinion big data and can predict the subsequent development of public opinion in the initial stage of the outbreak of public opinion,a parameter inversion algorithm of the public opinion propagation model using neural network is proposed based on the practical opinion big data on the basis of analyzing many public opinion forecasting models,such as neural network model,Markov model,etc.Finally,the applicability and effectiveness of the proposed algorithm are verified by inverting the data of "Wei Zexi" event and "man-machine war" event grasped on Sina micro-blog.The proposed algorithm can accurately predict the specific heat values of public opinion compared with the Markov prediction model and can be used for the data fitting,the process simulation and the trend prediction of the network emergencies' spreading.