| 摘要 |
<P>PROBLEM TO BE SOLVED: To provide a language converter for surely converting an inputted voice sentence into an objective language even when an unlearned part exists in the inputted voice sentence and a mistake is made in a part of voice recognition, and further to automatically prepare a phrase dictionary required for conversion and rules between the phrases without any manual work to the utmost, and also to provide a program recording medium. <P>SOLUTION: The device is provided with: storing means 3, 4 for storing a language rule obtained by learning grammatical or semantic restriction rules for words or word strings from a bilingual corpus where an original language sentence makes a pair with the objective language; a voice recognition part 7 for the inputted voice by using the stored language rule, and outputting a recognition result by the sentence to be the object of the language conversion; and a language conversion part 9 for converting the sentence to be the object of the language conversion into the language-converted sentence by using the language rule being the same as that used in the voice recognition part 7. <P>COPYRIGHT: (C)2004,JPO |
