Design And Implementation Of Network Public Opinion Monitoring System For Guangdong Police

The Internet is at hand, it gives people unlimited right to speak, changing people’s way of life. More and more Internet users express their views about social and public affairs, political attitudes, emotions, ideas, by the network. It has much explosive power and influence of public opinion, so that any institutions or individuals can’t ignore the point. Internet public opinion is a popular view about social issues on the Internet. It includes some influential opinions of the public about some hot spots in their lives.Analyzing information on the network will help to guide public opinion, so as to continuously maintain social stability. Strengthening the regulatory information of Internet public opinions will be good for correctly guide public opinion, so as to continuously maintain social stability and harmony. It is a very important work that the public security police must focus in the new era under the complex environment. And effective network monitoring technology of public opinion will largely reduce the strength of the grassroots work of police officers and increase working efficiency.Public opinion is a direct reflection of social public opinion on the Internet. The main carrier of the public opinion: news commentary, BBS, blog and micro blogging,RSS, news thread and so on. Internet Public Opinion for the Basic Public Security System’s aim is to warn and discover crisis events, crisis ahead of perception, especially,to properly control the negative public opinion in a timely manner, so as to achieve to resolve the crisis. The significance of public opinion monitoring network is to detect signs of the crisis in time, to judge the trend and scale, to inform the relevant functional departments to make preparations to tackle the crisis.In this thesis, it based on analyzing several key technologies about dealing with public opinion, they are information collecting technology, natural language processing, Web crawlers, and data mining technology, designing and implementing a network monitoring system of public opinion under the reality of working needs of grass-roots policemen. The system includes three main modules, information collecting of public opinion, information processing of public opinion, information analyzing of public opinion. In the system, using the DOM tree and improved PageRank model to process pages, laying a good foundation in data for text processing. In text processing stage, in order to effectively improve the efficiency and accuracy of text processing, Iuses TFIDF method to calculate the weights of feature words, using decision tree to classified documents, using the method of class cohesion level to cluster texts. Finally,the system will get a network of public opinion trends over time by analyzing the developing trend of information and Sentiment. The results of the experiment verify the thesis’ s research and the effectiveness of designing of the system.