Research On The Topic Mining And Emotional Trend Analysis Of Public Demands Under The Normalization Of The Epidemi

The COVID-19 epidemic has brought us not only huge economic losses,but also great psychological impacts and even changes some of our lifestyles and habits because of its sudden,harmful,long-term and other attributes.In the new situation of the normalization of the epidemic situation,there have been new changes in the theme and emotional situation of the people.In order to explore the evolution trend and theme classification of these changes,this article conducts research from both macro and micro perspectives.Studying the evolution trend of the theme of people’s appeals and the changing trend of people’s emotional situation from a macro perspective can help us grasp the evolution rule of people’s appeals and emotions under the new situation of normalization of the epidemic situation,and provide certain data reference for the macro control of social resources and public opinion guidance in similar public health emergencies in the future;Exploring the popular appeals and thematic vocabulary that lead to emotional changes from a micro perspective can help us understand the specific appeals of the people and provide accurate assistance and solutions.It also helps the official media guide public opinion in a positive way.Based on this,this paper takes the change of the public’s appeal hot spots and emotional situation under the normalization of the epidemic as the research entry point,uses the octopus data collector to obtain the data related to the epidemic on the message board of the local government of People’s Daily from January 2021 to June2022,and uses the Python web crawler to obtain the epidemic hot events reported on the official microblog of CCTV News and the comments of netizens under the article of the epidemic quick report during this period.The LDA theme model is used to carry out theme mining on the text data of the local government message board of People’s Daily Online.The popular appeal hot spots are divided into 13 theme areas and 38 sub-theme areas.At the same time,the social network analysis method is used to carry out the correlation analysis of the hot topic words,and the visual map of the correlation between the hot topic words is obtained.The analysis of the popular appeal hot spots from the micro and specific level is convenient for more targeted solutions.Then,according to the similarity and overlap of the themes,it is concluded that the appeal hotspots of the people under the normalization of the epidemic situation are divided into five theme types: economic policy,social security,infrastructure issues,leisure and entertainment,education and employment.Machine learning is used to analyze and study the emotional situation of the obtained Weibo comment text.Through accuracy testing,SVM is selected to judge the emotional polarity of the corpus.LDA topic models are used to categorize the hot events that lead to different emotions.Word Cloud word clouds are used to conduct word frequency statistics on emotional peaks,and topic words with different emotional peaks are obtained.This study describes the changes in the types of popular appeal themes and emotional situations under the normalization of the epidemic situation,as well as hot appeal themes and emotional hot words,which can provide certain data reference for the treatment of similar public health emergencies in the future.The research conclusions have certain practical significance.