Intelligent conversation support system based on discussion map and multi-modal interpretation

2016 Fiscal Year Final Research Report

• Project/Area Number: 26730176
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Learning support system
• Research Institution: Kyushu Institute of Technology
• Principal Investigator: Shimada Kazutaka 九州工業大学, 大学院情報工学研究院, 准教授 (50346863)
• Project Period (FY): 2014-04-01 – 2017-03-31
• Keywords: 自然言語処理 / 要約 / マルチモーダル / コーパス / 議論支援

Outline of Final Research Achievements

Supporting consensus-building on multi-party conversations is one of the most important tasks in intelligent systems. Participants on discussion often struggle to identify the most suitable solution for an agenda on a meeting because there are generally many alternatives and criteria related to the agenda. Such a supporting system for consensus-building has become a very important role for education in recent years because problem-based learning (PBL) is the more recent and highly regarded. In this research, I analyze discussions and proposed a tool for supporting consensus-building.
This project handled two tasks; (1) conversation summarization based on linguistic and non-linguistic features and (2) the discussion map system for consensus-building. In addition, I opened the Kyutech corpus, a freely available Japanese conversation corpus for a decision-making task, on the web.

Free Research Field

知能情報学