| 摘要 |
PROBLEM TO BE SOLVED: To make it possible to add interest to a speech response result or to provide information which a user does not know by outputting previously determined associative words in speeches in response with the recognition result by a speech recognition means. SOLUTION: A speech synthesis section 23 synthesizes the output of a word collation section 17, the output of an associative word reading out section 22 and the output of an application section 12 and outputs the same as speeches. If, for example, the application to be executed is dialing and when the user inputs 'weather forecast' with a speech, the speech inputted from a microphone 11 is recognized as the word 'weather forecast' by the speech recognition section 10. The associative word 'not trustworthy' which corresponds to the word 'weather forecast' is read out from an associative word dictionary section 21 and is synthesized to the speech 'the weather forecast is not trustworthy' by the speech synthesis section 23. The interest is thus added to the inorganic application of listening to the weather forecast over the phone. |
