Research On The Interactive Content Of Local Government Message Board Of People’s Daily Website Based On Text Mining

In recent years,benefiting from the vigorous development of China’s network infrastructure,the emerging network information technology represented by big data,cloud computing and artificial intelligence has risen rapidly,which has a subversive impact on people’s life style,especially in the field of interaction between the government and the people,the emergence of new media enriches the communication channels between the people and the government,enhances the willingness of the people to communicate with the government,and promotes the development of the communication mechanism between the government and the people in cyberspace.The government and the public can communicate with each other through microblogs,public accounts and online political platforms,which greatly improves public participation and promotes the interaction between the government and the people.However,it is undeniable that in today’s e-government era,for massive government data,how to use big data technology to mine valuable information,improve the efficiency of government-civilian interaction,increase people’s satisfaction,and strengthen government credibility for massive government data has become a pressing issue in government work.The issues to be solved have important practical significance for improving the government-civilian exchange mechanism and promoting the development of our country’s e-government to a deeper level.Based on the existing research results in the field of text mining and network interaction between government and people,this paper uses the theme extraction technology,emotional analysis method and hypothesis testing statistical method,combined with the theory of citizen participation,public crisis management theory and e-democracy theory to study the people’s message data of Sichuan Province on the local government message board of people’s network,explore the hot topics and emotional tendencies of the people in the messages,and analyze the differential influence of the government’s response speed in different dimensions.First of all,this paper uses LDA topic model to extract the topic of public comments,mining the hot topics of public concern,and taking the time dimension into account,it is found that there is a certain fluctuation trend of hot topics in different years.Secondly,the sentiment dictionary method is used to analyze the sentiment of the message data.Combined with the LDA topic extraction results,the sentiment tendency of different topics is analyzed.The results show that there are more negative sentiment messages on topics such as living environment,and more positive sentiment messages on topics such as poverty alleviation project.Finally,for the research on the difference of government response speed,this paper uses Kruskal Wallis nonparametric test method,and finds that the theme,region and emotional tendency of public messages have different effects on government response speed.The research of this paper broadens the research ideas in the field of the interaction between the government and the people,helps the government to grasp the hot spots and psychological situation of the people in time,and improves the efficiency of the interaction between the government and the people.