| Internet Public Opinion Security Warning System is a kind of software that can discover the network of potential network security risks and achieve real-time monitoring of networks. This software can analysis based on facts and experience to judge for the hotspots, points, frequency points, sensitivity points, doubts on the web based on management science, statistics, ontology, and behavioral science and computer technology, simulation technology and information technology as a kind of manner to work.Chinese word segmentation algorithm is now maturing. Including Largest Positive Matching algorithm, Maximum Reverse Matching algorithm, Two-way Matching algorithm and so on. The system uses a combination of multiple algorithms approach to achieve the intelligent word search function based on the information entered by the user with specific rules.In the thesis, the author first introduces the research background, situation at home and abroad, and the tasks of the project. Then, the author introduces the implementation of the three modules that "Public Opinion", "Public Opinion Search", "User Behavior Tracking", J2 EE, Weka Data Access and Java GUI. According to requirements, the author improves algorithm, presents architectural and detailed design for each functional module. The author has taken a lot of experiments and presents the data and result in the paper. The author also discusses the algorithm and interface design and implementation demonstration at last. |