A Novel Method of Topic Detection and Tracking for BBS
Topic Detection and Tracking (TDT) has been studied for years, but most existing research is oriented to news web pages. Compared to news web pages, texts in Bulletin Board System (BBS) are more complicated and filled with user participation. In this paper, we propose a novel method of TDT for BBS, which mainly includes: a representation posts selection procedure based on post quality ranking and an efficient topic clustering algorithm based on candidate topic set Experiment results demonstrate that our method significantly improves the performance of TDT in BBS environment on both accuracy and time complexity.
topic detection and tracking BBS representation posts selection candidate topic set clustering
Yan Zhao Jungang Xu
School of Information Science and Engineering Graduate University of Chinese Academy of Sciences Beijing, China
国际会议
西安
英文
453-457
2011-05-13（万方平台首次上网日期，不代表论文的发表时间）