摘要 |
PROBLEM TO BE SOLVED: To provide a voice recognition dictionary-editing device and a voice recognition device, which can efficiently edit voice recognition data of a voice recognition dictionary even when the editing device has no function of editing a portion of voice recognition data registered in the voice recognition dictionary accumulating voice recognition data registering text and pronunciation data of the text about each of texts to be recognized. SOLUTION: A voice recognition dictionary is created for each of titles of n pieces of music (901). If the reading of a music title is changed by a user, a reading conversion rule of the music title is revised. The voice recognition dictionary containing reading data of the music title of the changed reading is created again by using the reading conversion rule after the revision (902-904). COPYRIGHT: (C)2011,JPO&INPIT
|