VOICE RECOGNIZING METHOD

摘要

PROBLEM TO BE SOLVED: To provide a method for recognizing a voice which is improved in recognition rate. SOLUTION: The method for recognizing a voice includes the steps of: receiving a user's speech input (SI); determining a arranged hypothesis (OH) group including tag information (TI) representing at least one kind or one modification of pronunciation for respective hypotheses in the arranged hypothesis (OH) group when the hypothesis (OH) group arranged for the received voice input (SI) is determined; using a tag language model (LM2) operating on the tag information (TI); rearranging the hypothesis group by using the tag language model (LM2); and outputting the rearranged hypothesis (ROH) group to select the best hypothesis (BH). COPYRIGHT: (C)2005,JPO&NCIPI