- 著者
-
今井 健太
北村 光司
西田 佳史
竹村 裕
杉本 裕
- 出版者
- 一般社団法人 日本機械学会
- 雑誌
- ロボティクス・メカトロニクス講演会講演概要集 2015 (ISSN:24243124)
- 巻号頁・発行日
- pp._1A1-W02_1-_1A1-W02_4, 2015-05-17 (Released:2017-06-19)
A large number of injuries were occurred in Japanese school environments. To prevent school injuries, it is important to understand accident situations and prioritize intervention targets. However, it is difficult for a risk manager of each school to do this since each school does not have data enough for statistical analysis. In this study, we developed the system that allows to grasp serious accident situations by integratively utilizing data distributed in multiple schools. The developed system finds serious accident situations that a school risk manager should know as follows. First, the system registers situational feature vectors for accident situations data by using a textmining technology. Second using the database, it searches accidents situations similar to the actually occurred situations. Finally it shows most expensive accidents using medical cost data. The effectiveness of the system was confirmed using 5,817 school injury data.