| 摘要 |
PROBLEM TO BE SOLVED: To make intonation of learner's speech closer to that of a model speech in language learning.SOLUTION: When one of displayed example sentences is selected, a language learning device reproduces model speech when the example sentence is uttered. Then, a first pitch curve for indicating time change of a pitch of the model speech is formed, and a route of the first pitch curve is calculated. The language learning device prompts the learner to utter the example sentence, when reproduction of the model speech is finished. When the learner's speech is input, the language learning device 1 stores the input speech. The language learning device 1 forms a second pitch curve for indicating time change of the stored learner's speech, and a route of the second pitch curve is calculated. Then, the language learning device compares the route of the first pitch curve with the route of the second first pitch curve, and according to difference of the routes, difference of the pitch of the model speech and the pitch of the learner's speech is output to the learner. |
