| Joseph Irwin | 1261023: D, 中間発表 | 松本 裕治, 中村 哲, 新保 仁, Kevin Duh |
|
title: Learning Scripts By Predicting Verbs
abstract: Many applications in NLP benefit from world knowledge. One type of world knowledge, narrative schema, can be learned from unannotated text. The purpose of this research is to develop an improved representation and algorithm for learning this kind of discourse-level knowledge. In order to facilitate this, I propose an intrinsic evaluation method -- a verb prediction task -- which is meant to enable easier comparison between similar but distinct representations. language of the presentation: English | ||
| 重藤 優太郎 | 1251050: M, 2回目発表 | 松本 裕治, 池田 和司, 新保 仁, Kevin Duh |
| 白井 尊昭 | 1251053: M, 2回目発表 | 松本 裕治, 池田 和司, 新保 仁, Kevin Duh |
| 小林 龍 | 1151048: M, 2回目発表 | 松本 裕治, 池田 和司, 新保 仁, Kevin Duh |
| 藤本 雄一郎 | 1261012: D, 中間発表 | 加藤 博一, 横矢 直和, 宮崎 純, 武富 貴史 |
|
title: Research for task support system by Augmented Reality
abstract: Augmented Reality (AR) is the technology that overlays virtual images onto the real world. The technology has been popular because of the applications for gaming and the smartphone. However, only few applications exist in the field of task support systems which has been the most important goal of AR. Taking into account the current situation, we organize exising AR task support systems and clarify the existing problems in it. In addition, we tackle two obstacles for applying AR to the task support system: "shortage of the knowlegde of its advantages" and "technical problems". As the former, we investigate the relationship between AR and the user's memorization to obtain new advantages of AR for task support. As for the latter, we try to consider technologies for the position and orientation estimation of "deformable objects" and "specular object" with some difficulties. language of the presentation: Japanese | ||
| ANGIE CHEN | 1251122: M, 2回目発表 | 加藤 博一, 横矢 直和, 武富 貴史, 山本 豪志朗 |
| 宮�� 良 | 1251104: M, 2回目発表 | 加藤 博一, 横矢 直和, 武富 貴史, 山本 豪志朗 |