Research on analysis technology of Internet Public Opinion based on topic cluster

Network has become the main vector of information dissemination, and, the information stored and transmitted in Internet can be reflected Internet Public Opinion in period. The Sensitive topic mining study on the Internet helps us understand the Internet public opinion trends. This paper studies method about search data cluster of Internet users with K-nearest neighbor and shortest path approaches. We first build an association search net, then, provide shortest path. It will help us analyze Internet user´s search behavior and characteristics. Finally, we will discover the information dissemination pattern and guide Internet public opinion trends correctly.