2019 Fiscal Year Annual Research Report
Development of a computerized speech assessment & feedback system
Project/Area Number |
17K02937
|
Research Institution | Kochi University of Technology |
Principal Investigator |
ダニエルズ ポール 高知工科大学, 共通教育教室, 教授 (50307245)
|
Co-Investigator(Kenkyū-buntansha) |
熊井 信弘 学習院大学, 付置研究所, 教授 (00248999)
|
Project Period (FY) |
2017-04-01 – 2020-03-31
|
Keywords | language / learning / speaking / speech / recognition / computer-scored |
Outline of Annual Research Achievements |
1. An open source automated speech scoring system was successfully developed over the past three years to administer practice speaking tasks within a university English language learning curriculum. In 2019, the auto-scored speaking tasks were assigned to a wider group of language learners with positive results. Data collected from the system confirmed that the computer generated speaking scores were consistent with popular propriety automated scoring systems. 2. After extensive testing with language instructors and learners, additional functions were added to improve the system. The improvements to the speech system are outlined here. a. The ability to choose a specific speech engine for speech-to-text support, for example Google or Amazon, was implemented. b. User feedback was improved to highlight phonemes, words, or phrases spoken by the learner that were not accurately recognized by the speech engine. c. The system was made compatible with mobile devices using Android and iOS operating systems. d. An automatic audio reply function was added to create more interactive speaking tasks.
|