|
|||||
|
|
Comments’ Extraction by User ID,Keywords, and Anchor Texts from the Thread of a Bulletin Board System |
|
| Authors: | Yoko Nishihara Wataru Sunayama Kazunori Nishimura |
| Institution: | 1. College of Information Sciences and Engineering, Ritsumeikan University, 1-1-1 Noji-higashi, Kusatsu, 525-8577, Shiga, Japan 2. Graduate School of Information Sciences, Hiroshima City University, 3-4-1 Ozuka-higashi, Asa-minami, 731-3194, Hiroshima, Japan 3. Faculty of Engineering, Hiroshima Institute of Technology, 2-1-1 Miyake, Saeki, 731-5193, Hiroshima, Japan |
| Abstract: | Bulletin Board Systems (BBSs) on the Web are used by many users posting comments on threads. Each thread carries a subject of discussion. Most users post comments related to a subject. However, unrelated comments are also posted. The unrelated comments distract users and diminish their understanding of the entire story of a thread. Comments unrelated to a subject have to be filtered automatically. This paper proposes a method that extracts comments related to a thread subject from a thread of BBS. The method extracts two types of comments: comments related to a thread subject (main comments), and comments related to main comments (sub-comments). The main comments are extracted by a user ID and keywords in the comments. The sub-comments are extracted by explicit and implicit anchor texts in the comments. We experimented with the proposed method and verified that comments extracted by the proposed method support participants in understanding the entire story of a thread. |
| Keywords: | |
| 本文献已被 SpringerLink 等数据库收录！ | |