Research On The Prediction Of Public Opinion On Public Health Emergencies In Terms Of Heat And Emotional Evolution

With the rapid development of the Internet and the emergence of various applied media platforms,the explosive growth of data brings a large amount of emotional information to public opinion.In addition,in view of the destructive and uncontrollable characteristics of public health emergencies,the spread of emotional tendencies will be faster,which is more likely to cause negative emotions such as people ’s anxiety and anxiety,and will seriously affect the stable order of society.Therefore,in different stages of public opinion,how to accurately and efficiently mine hot topics from complex information for public health emergencies,predict the trend of heat development,and grasp the emotional trend of various public opinion information and topics is of great significance to the government and other public opinion control departments in order to guide and control the developmental tendency of public opinion,maintaining a healthy and clear networked environment,and ensuring social harmony and stability.Given the research content of this article,divides primarily into the following three sections:(1)Research on public opinion topic extraction of public health emergencies.This thesis adopts the method based on LDA topic model to extract the topic of Weibo comments on public health emergencies based on life cycle theory from the perspective of text.Then,through the case analysis of the crawled ’ COVID-19 ’ comment data,through data preprocessing and other operations,the keywords and topics of people ’s comments at different stages are analyzed,so as to grasp the main content of people ’s attention at each stage of the event and excavate its hidden hot topics.(2)Research on public opinion heat prediction of public health emergencies.From the perspective of non-text,the whale algorithm is constructed to optimize the Elman neural network heat prediction model,and the evolution law of heat in the process of public opinion information dissemination is excavated.By collecting the heat value of the comprehensive time series and comparing it with the standard BP and Elman neural networks with the help of relevant evaluation indicators,the prediction accuracy and prediction error of the model are analyzed to verify the accuracy and effectiveness of the model in heat prediction.(3)Research on emotional evolution of public opinion in public health emergencies.In view of the changes in emotional attitudes brought about by the evolution of public opinion,this thesis proposes an emotional evolution model based on naive Bayes from the perspective of emotion.Combining the results of sentiment analysis with the results of topic extraction in Chapter 3,this thesis studies the changes in content and emotion of the public opinion event,and provides corresponding countermeasures for the public opinion control department at each stage by analyzing the emotional differences at different stages.According to the characteristics of public opinion of public health emergencies,this thesis studies the text theme and the heat of public opinion at each stage from the perspective of text and non-text,and based on the above analysis,introduces emotional factors to conduct the emotional analysis of the text comments and the extracted themes.The results show that the model proposed in this thesis has good results in theme extraction,heat prediction and emotion analysis,which enriches the research content of public opinion,and also provides new ideas for the governance and control of relevant departments,and has certain reference value.