摘要 |
<p>PROBLEM TO BE SOLVED: To reduce an error generating rate for speech speaking spelling-read at least partially. SOLUTION: In this automatic recognizing method, a speech recognizing unit 2 determines the first recognition result relating to speech speaking after the speech speaking is input, characters individually recognized are transmitted to a user for user's approval or rejection, a language speech model 6 is adopted when the character is approved by the user, the model 6 after adopted determines the number of characters to be followed to the approved character, the already approved character is determined to be accurate, the speech recognizing unit 2 determines a further recognition result relating to the speech speaking using the adopted language speech model 6, a next character transmitted to the user is determined to be approved by the user based on the further recognition result.</p> |