Verbalization of data analysis or future prediction

2016 Fiscal Year Final Research Report

• Project/Area Number: 26540190
• Research Category: Grant-in-Aid for Challenging Exploratory Research
• Allocation Type: Multi-year Fund
• Research Field: Entertainment and game informatics 1
• Research Institution: Kyoto University
• Principal Investigator: Mori Shinsuke 京都大学, 学術情報メディアセンター, 教授 (90456773)
• Project Period (FY): 2014-04-01 – 2017-03-31
• Keywords: ゲーム / 言語生成 / 情報検索 / 自動解説 / シンボルグラウンディング

Outline of Final Research Achievements

We have proposed and realized a method for generating commentaries for a given game state and possible future states.
We have built a game corpus through this research and presented it in LREC 2016. With this corpus we have developed a method for searching game states by natural language keywords and made a presentation in the top conference of information retreivel, ACM SIG IR. We have also shonw that real world information improves the accuracy of a named entity recognizer. We made a presentation on this research result in one of the top conferences of natural language processing, ACL 2016.

Free Research Field

自然言語処理