2018 Fiscal Year Final Research Report
Research and development of an automatic judging system for determining the difficulty levels of English audio materials
| Project/Area Number |
15K12420
|
|---|
| Research Category |
Grant-in-Aid for Challenging Exploratory Research
|
| Allocation Type | Multi-year Fund |
|---|
| Research Field |
Educational technology
|
|---|
| Research Institution | Akita Prefectural University |
|---|
Principal Investigator |
OKAZAKI Hironobu 秋田県立大学, 総合科学教育研究センター, 教授 (80405084)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
渡辺 貫治 秋田県立大学, システム科学技術学部, 准教授 (20452998)
福田 衣里 中国学園大学, 公私立大学の部局等, 講師(移行) (50617488)
橋本 信一 電気通信大学, 情報理工学域, 特任准教授 (60350500)
木戸 和彦 環太平洋大学, 次世代教育学部, 准教授 (80599184)
|
|---|
| Project Period (FY) |
2015-04-01 – 2019-03-31
|
| Keywords | e-ラーニング |
|---|
| Outline of Final Research Achievements |
In order for a learner-selected news audio file to be used as English listening material, the difficulty level of the audio selection must first be determined. Then, its appropriateness as a listening material for a particular learner can be assessed. In this study, we aim to develop an automatic judging system that determines the difficulty levels of audio samples by using an acoustic analysis approach. This approach utilizes a large database of audio samples which was gathered previously.
To achieve this objective: 1. A diagnostic test composed of 180 sentences of mostly English words was given to Japanese learners to collect data on listening mistake tendencies. 2. A website to collect data on the difficulty levels of English learning materials for international students was constructed. 3. A model system for automatic judging was created and a recognition engine for the model was selected. 4. A voice-recognition engine which differs from current approaches was newly developed.
|
| Free Research Field |
教育工学
|
| Academic Significance and Societal Importance of the Research Achievements |
従来のレベル別教材は、あらかじめ用意された教材、つまりレディメイド教材であり、学習者の細かいニーズを満たすものではなかった。しかし、本研究により、任意の音声素材のリスニング難易度を自動判定するシステムが構築され、e-ラーニング用プログラムに組み込めれば、学習者のレベルだけではなく、趣味や嗜好といったモチベーションにかかわる部分までもカバーするオーダーメイド音声教材の自動提示が可能となるのである。
|
