Font Size: a A A

Research On Public Opinion Analysis Methods For Specific Events

Posted on:2019-05-20Degree:MasterType:Thesis
Country:ChinaCandidate:H ShaoFull Text:PDF
GTID:2428330566984177Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
Network public opinion is a form of expression of public opinion.It has strong influence and tendency on some hot spots and focus problems in real life through the Internet.In recent years,the influence of network public opinion on the order of political life and social stability is increasing day by day.Some important network public opinions have made people realize the great role of network on social supervision.At the same time,if the network public opinion emergencies are not handled properly,it is very likely to induce the bad mood of the people,cause the people's irregularities and excesses,and then pose a threat to the social stability.Therefore,it is of great value and significance to study the trend of public opinion.Especially in recent years,the violence in Xinjiang has become a focus.Then,it is very important to grasp the dynamics of public opinion in Xinjiang's violence.Evolution is an important part of the public opinion analysis.In order to grasp the trends of public opinion in Xinjiang,we use evolution methods to do research on the key words of topic and the change of both topic strength and its content.Current topic evolution methods are mainly divided into probabilistic models based on LDA and matrix decomposition models.NMF is a kind of matrix decomposition which is suitable when the task is to identify both broad,high-level groups of documents and niche topics with specialized vocabularies.Considering that the violent terrorist incident in Xinjiang itself has a big theme of violence,this paper proposes an improved method based on NMF,and we get more coherent topic descriptors by changing the weights of words.In this model,two times of NMF are used to generate the topic,and then the hierarchical Dirichlet process is used as a contrast experiment.Through visual analysis and comparison,some rules of Xinjiang's violent events are summarized.Social platform is an important way to spread public opinion,and the information of the influential media can be spread faster and wider.The analysis of media influence can clarify the degree of information dissemination and grasp the public opinion in some degree.On the other hand,we use the improved PageRank method with media's individuation to analyze the influence of media about Xinjiang Violence on the twitter,and then shows the attitude and influence of various countries' media which providing a reference to grasp the development of the public opinion of Xinjiang Violence.
Keywords/Search Tags:Internet public opinion, Topic Model, NMF, Xinjiang Violence, PageRank, Visualization
PDF Full Text Request
Related items