<![CDATA[Network public opinion analysis is an important way of information analysis processing. This paper based on the research of the related technologies, designs and realizes a new network public opinion analysis system. System mainly includes network data fetching part, fetching the data processing part, analyzes the processed data part and display part of the public opinion analysis results. In the document extraction part, used the web crawler technology, Larbin web crawler to realize the collection of web content; In public opinion information analysis part, the implementation of the new topic adopts an improved Single - Pass clustering algorithm. This algorithm is using of multi-center, using the title and body of the vector to compared two-way ,that is better reflect the dynamics of public opinion topics. Finally, in the network environment of a university, we have the tests repeatedly. The results show that the new public opinion analysis system running is stable and has good efficiency. The thesis has certain value for the development of other information analysis systems in the Internet.]]>
Copyrights © 2015
