Automatic Extraction of Task Statements from Structured Meeting Content

• DOI: 10.5220/0005609703070315
• Publication Date: 2015-12-02T11:41:40Z
• Authors: Katashi Nagao; Kei Inoue; Naoya Morita; Shigeki Matsubara
• Keywords: Statement (logic)

We previously developed a discussion mining system that records face-to-face meetings in detail, analyzes their content, and conducts knowledge discovery. Looking back on past content by browsing documents, such as minutes, is an effective means for conducting future activities. In at which some research topics are regularly discussed, seminars laboratories, the presenters required to discuss issues checking urgent matters from records. call statements including advice or requests proposed previous “task statements” propose method automatically extracting them. With this method, based certain semantic attributes linguistic characteristics of statements, probabilistic model created using maximum entropy method. A statement judged whether it task according its probability. seminar-based experiment validated effectiveness extraction