The fingerprint analysis technique-oriented research on microblog for public opinion analysis

Among all the Internet applications, microblog is warmly welcomed by the users for its concise content, flexible form, rapid spread and real-time update. The total number of microblogs registered in China's main platforms, such as Sina, Tencent, Netease, Sohu and so on, exceeds 1 billion. Because of the huge number of the user, microblog seems to have become an independent route of information transmission, which can work independently in the guidance of public opinion. How to regulate microblog and locate public opinion hazards, and make use of microblog for better planning and guiding the public opinion, and create elegant and pleasant cultural life should be our new concern. In this paper, according to the characteristics of microblog content and form, we propose a technology based on the analysis of the fingerprint information. We use network spiders crawling microblog content with limited public opinion keywords, then extract the text content and use simhash algorithm creating the unique fingerprint. Fingerprints are clusterd base on hamming distance. In the same category, we analyze the data, discovering the sensitive topic, looking for key information disseminators, thus control the public opinion. After many times of experiments, this method has a higher accuracy and credibility.