摘要 |
PROBLEM TO BE SOLVED: To solve the following problem; with a conventional language training device, a learner of a language, mimicry, or the like cannot grasp a degree of similarity between model speech to learn and speech uttered by the learner while performing what is called follow-up reciting (also called "shadowing"). SOLUTION: While performing follow-up reciting with the pronunciation grading device comprising; a teacher data storage part storing a teacher data which is a data concerning speech for comparison; a speech output part for outputting speech corresponding to the teacher data; a speech receiving part for receiving the speech input after the speech output part outputs at least a part of the speech data; a grading part for grading the speech received by the speech receiving part based on the speech received by the speech receiving part and the teacher data; and an output part for outputting a result of the grading by the grading part, a degree of similarity between the model speech to learn and the speech pronounced by the learner. COPYRIGHT: (C)2006,JPO&NCIPI
|