Design And Implementation Of The System Of Public Opinion Analysis Based On Microblog

With the rapid development of Internet, more and more people are willing to express their thoughts, emotions and attitudes through the Internet. As one of the representatives of new media, microblog has gradually become the information sharing, communication and information access platform. Every day, there are millions of microbloggers on the Internet, which are timely and full of variable information.It becomes more important and difficult how to mine and analyze the main content from the microblog data. In this thesis, we design the public opinion analysis system based on Sina microblog. First, the collection and pre-processing of microblog data is proposed, and then the text classification for microblog data is also present. Finally, this thesis presents the approach on text clustering. The main contributions are shown as follows:1) The thesis presents the approach on microblog data collection and pre-processing. On the basis of the subject-oriented Sina microblog data crawling, the thesis presents the method on microblog data collection, and the crawled data is stored in the database. After that, some pre-processing tools are used to clean the above data, such as the processing of Chinese text, Chinese word segmentation, and word frequency matrix dimensionality reduction.2) We analyze the performance on the subject-oriented classification on the basis of the microblog data pre-processed. Through the Chinese text automatic classification algorithm to model and determine the content of each microblog theme, the experiment uses make use of the current popular classification algorithm(KNN, decision tree, random forest) to evaluate and analyze the effect of the microblog data。3) The thesis presents the clustering Analysis of micro blog text clustering algorithm based on K-Means algorithmThe experiment proves the performance of the algorithm.The result shows that the research of the microblog text analysis algorithm has some value in scientific research and social application field. In the end, this thesis also presents the existing problems and the further research works.