Microblog Public Opinion Topic Evolution Model Based On Wikipedia Semantic Clutering

Microblog is an important source of data for public opinion analysis systems.Microblog public opinion topic evolution can extract the public opinion topic from microblog text flow and find the trend of public opinion over time,and help users quickly and accurately understand the context of public opinion and track topic evolution,which has a very important significance for the government to conduct public opinion and the business to mine business information.Combined with the characteristics of microblog text and the law of microblog public opinion topic evolution,this paper proposed an evolutionary model of microblog public opinion topic based on Wikipedia semantic clustering.This model descript the evolution of public opinion topic from two aspects of the public opinion:content and strength.Firstly,against the question studied in this paper,we studied and summarized the current research situation and made a detailed study on the theoretical basis and technical basis of the proposed model,which purpose is propose the research method on the basis of the existing research.This paper mainly made an in-depth study and research on the text clustering technology,the semantic application technology based on Wikipedia and topic evolution technology.Secondly,topic extraction is the premise of public opinion topic evolution,this paper proposed a method based on Wikipedia to extract and express the public opinion topic.This method use Wikipedia as the semantic knowledge base of microblog text feature words,and use Wikipedia link structure and articles to improve the K-means semantic clustering algorithm,which used to complete the clustering on microblog text of different time slices,and describe the public opinion topics by important feature words and microblog text of the cluster.Thirdly,we calculated the similarity between topics to describe the evolution relationship between the theme of the adjacent time slices,then drew chart to show public opinion theme evolution;we also proposed a measurement method of topic intensity based on microblog heat,which is used to describe the intensity evolution of public opinion,then drew chart to show the evolution of topic intensity.Finally,the empirical analysis is carried out on the microblog text dataset to verify the validity of the model proposed in this paper.