Research Of Detection And Tracking On The Chinese Micro-blog Topic And System Design

Micro-blog as new media rise above the common herd of a Web2.0in the informationage，multimedia platform to support cross-platform information interaction，is developingrapidly in recent two years，has gradually become the main platform for ordinary people toshare personal information，pay attention to others’ information，real-time informationacquisition? has gradually become the main part of the network media. Its characteristic is thehuge number of information, decentralized，diversity.In order to let the user real-time understanding of the overall topic of micro-blog，tracking their interest in the topic，this paper Chinese micro-blog topic data acquisition，tracking method of topic detection.Through the use of suitable micro-blog the webpageinformation collection technology一time control breadth-ifrst acquisition based oninformation collection，improve eiffciency，ensure the information acquisition coverage.Adaptive collection and information extraction of micro-blog site topic identiifcation andstandardized，modular storage，to provide better quality of the data source.At the same time of obtaining micro-blog based on API data，and compared the webcrawler data acquisition mode based on API and obtain micro-blog data based on mode twoschemes in micro-blog data acquisition performance.By the end of the Chinese processing techniques for text processing， detection andtracking algorithm is used to obtain the data. In topic tracking real-time adjustment of queryvector process，and by introducing the webpage relationship，core features and non corefeature adjustment effectively iflter the noise information，and enhance the query vectoradjustment effect. Ultimately the micro-blog topic detection and topic tracking.