The rumors spread faster than the truth through online social media, and online discussions can also be affected by those rumors and misinformation. Thus, it is necessary to design a system to detect rumors and information during online discussions. The past research focusing on rumor detection has some issues: 1) Most of the fact-checking works are based on a single source which is assumed to be authoritative. 2) Checking if the information source is authoritative is an endless loop. 3) Multiple trustworthy information sources may conflict with each other, and only one ranking standard could be established in the past. 4) Binary label classification is not suitable for posts of online discussions. Therefore, a multi-agent system framework connected with LLM is conducted to mutually verify if posts of online discussions are trustworthy.
噂・不正言論の検出。
| 氏名 | コース | 研究室 | 役職/学年 |
|---|---|---|---|
| Dong Yihan | 社会情報学コース | 伊藤研究室 | 博士2回生 |
| 伊藤 孝行 | 社会情報学コース | 伊藤研究室 | 教授 |