表題番号:2019R-034
日付:2020/04/10
研究課題Modeling speech fluency for open-source e-learning application development
研究者所属(当時) | 資格 | 氏名 | |
---|---|---|---|
(代表者) | 理工学術院 創造理工学部 | 准教授 | ローズ ラルフ レオン |
(連携研究者) | Waseda University | Professor | Sheppard, Chris |
(連携研究者) | Waseda University | Associate Professor | Kinoshita, Naoko |
(連携研究者) | Waseda University | Assistant Professor | Shinohara, Yasuaki |
(連携研究者) | Waseda University | Professor | Sagisaka, Yoshiori |
(連携研究者) | Waseda University | Professor | Kondo, Mariko |
(連携研究者) | Waseda University | Doctoral student | Sun, Yue |
- 研究成果概要
- This project aims to develop a shareable model of second language speech fluency that can be utilized by both language teaching practitioners as well as language learning technology engineers and programmers. This year’s efforts were focused on two main goals: first, continued development and validation of a speaking practice application called Fluidity that embodies such a model; and second, the organization of an exchange of ideas regarding fluency modeling for theoretical and practical purposes. On the first goal, the Fluidity application detection mechanisms were subjected to some acoustic validation tests to gauge how accurately the speech, pause, and syllable detection mechanisms work. The results of these tests were quite positive, showing that Fluidity performs better than a widely-used script based on the Praat acoustic phonetics analysis tool. These results are to be published in the proceedings of the annual Speech Prosody conference. On the second goal, informal discussions with the research collaborators generated interesting ideas about fluency modeling including a debate about how to make such modeling useful to such diverse fields as pedagogy and engineering. Unfortunately, a plan to host a public meeting to discuss these topics could not be carried out due to the COVID-19 situation.