摘要 |
PROBLEM TO BE SOLVED: To provide an apparatus for automatically generating punctuation in a continuous voice recognition and a method therefor. SOLUTION: This apparatus automatically generates the punctuation with a continuous voice recognition system including means 1, 2, 3, 5 for recognizing user's voice and converting the user's voice into words. In such a case, the means 1, 2, 3, 5 for recognizing the user's voice are used also for recognizing the pseudo noise in the user's voice. Further, the apparatus includes a means 9 for marking the pseudo noise in the output results of the means 1, 2, 3, 5 for recognizing the user's voice and means 10, 14, 13 for generating the punctuation by finding out the pseudo punctuation of the highest probability in the position of the pseudo noise marked by the means 9 for marking the pseudo noise in accordance with the language model including the pseudo punctuation.
|