2006 Fiscal Year Final Research Report Summary
Research of Synthetic Speech Effectiveness in Foreign Language Education
| Project/Area Number |
17500640
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Educational technology
|
|---|
| Research Institution | Kobe University |
|---|
Principal Investigator |
KANG Min Kobe University, Faculty Of Cross-Cultural Studies, Associate Professor, 国際文化学部, 助教授 (60290425)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
KASHIWAGI Harumi Kobe University, School Of Languages And Communication, Associate Professor, 国際コミュニケーションセンター, 助教授 (60343349)
OHTSUKI Kazuhiro Kobe University, Faculty Of Cross-Cultural Studies, Professor, 国際文化学部, 教授 (10185324)
KABURAGI Makoto Kobe University, Faculty Of Cross-Cultural Studies, Professor, 国際文化学部, 教授 (40093504)
|
|---|
| Project Period (FY) |
2005 – 2006
|
| Keywords | Listening Skills / Text-To-Speech / CALL / Foreign Language Education / Materials Development / Evaluation |
|---|
| Research Abstract |
We investigated effectiveness of synthetic speeches in listening learning in order to find out if synthetic speeches are suitable for foreign language learning, what features of the sounds that can be utilized in listening materials are and how synthetic speeches influence learners' listening perception. Surveys of speech evaluation both for learners and teachers at the word and sentence level are conducted and listening training experiments are performed. In the surveys, natural voices were also used and the words were classified into three groups called "general words", "words with fricative" and "words based on vowel classification" according to their sound features. The results of the surveys reveal that learners show little difference on comprehending parts of the synthetic voices in comparison with the natural voices. The voices are those with the short vowels /Ι/, /υ/ and /*/, the long vowels /i:/, /u:/ and /*:r/, the diphthongs /aΙ/, /*Ι/. For "words with fricative", the evalua
… More
tions correspond to types of fricatives while synthetic voices at the sentence level are easier to understand and acceptable in comparison with synthetic voices at the word level. The evaluations by teachers exhibit that there is no significant difference between synthetic voices and natural voices although the evaluations on some voices are in agreement with those of learners. The listening training experiment consisted of pretest (natural voices), listening training (synthetic voices) and posttest (natural voices) and performed for English and Chinese respectively. In both experiments, learners showed the improvement in listening comprehension. We suggest that synthetic speeches are suitable to be selectively used in foreign language learning and listening training with synthetic speeches is efficient to enhance listening skills of learners. In addition, in order to find out adequate methods using synthetic voices in classrooms, we investigated the effectiveness of communication tools and RFID. Less
|
Research Products
(12 results)
