| 摘要 |
PROBLEM TO BE SOLVED: To provide a code recognizing device capable of recognizing the voice including surplus information, without fail. SOLUTION: The code recognizing device urges the utterance of a former half phrase of postal code number in United Kingdom (S210) and recognizes the voice (S220). If the result is "CV2", this device specifies the number section of a latter half phrase being connectable with "CV2" (S230). The device creates a dummy recognition dictionary by enumerating all sorts of number sections for the specified number section of the latter half phrase (S240), then urges utterance of the latter half phrase (S250), and recognizes the voice input through a voice inputting part 13, by collating it against the prepared phrase (S260). If a user utters "2TB", this device recognizes it as "2TB" and recognizes the postal code number, based on the number sections in the recognized former half phrase and latter half phrase (S270). If the former half phrase is CV2 and the latter half phrase is 2TB, the device recognizes them as being "CV2-2". COPYRIGHT: (C)2010,JPO&INPIT |
