Research Of Hot-topic Discovery About Internet Public Opinion

With the development of science and technology, as a new form of media, Internet has been playing a decisive position in fierce competition of the media. It is inevitable that the Internet public opinion has influenced the society profoundly. In regard to the Internet public opinion, their occurrence range is wide, spreading speed is high, and their eruption spot is difficult to be detected and controlled .Therefore, not only governments but also individuals are necessary to master the hot-topic timely, know the direction of popular opinions correctly, and take action to control and guide the trend of hot-topic in order to reduce the disadvantageous influence.BBS has become a popular communication tool at present. Through the BBS, Internet users could launch a new topic, reply to existing topics, exchange information or express ideas of topics. To some extent, BBS is a mirror of real society. At the same time, the real-time data on BBS can be collected which provides certain data base for the research of the Internet public opinion. In this thesis, analysis and monitoring function are based on real-time data of BBS.Based on the analysis of the domestic and international development and status of Internet hot-topic detect research, we deeply analyze how to design an effective architecture of the Internet public opinion analysis system and what is the proper system process in this thesis. The Internet public opinion analysis system is developed on Visual Studio 2005 platform which is a tool of Integrated Development Environment. The background database of the system is Oracle9i which is a kind of relational database. The C/S (Clint/Server) system development structure is adopted in the system. Multithread, MD5 encryption and other technologies are adopted in implementation of the system. In this thesis, we will introduce main function, system process and technique used in implementation of some key assistant modules which names are Internet public opinion information collecting module and system management module. The information collecting module will provide data source for hot-topic analysis and monitoring. Automatic removal of duplication is introduced to assure the accuracy of the real-time data.In this thesis, the function of hot-topic analysis and monitoring for BBS are implemented. A topic activity model based on the quantity and heat level of topics is adopted in hot-topic analysis and monitoring. Though measuring topic activities of all the topics on BBS, hot-topics will be quickly filtered according to time coordinate .The function of hot-topic analysis and monitoring can find hot-topics according to current information instead of historical information, thus the efficiency of finding hot-topic is improved and the time complexity is greatly reduced.