| 摘要 |
<p><P>PROBLEM TO BE SOLVED: To realize proper guidance by covering the mis-recognition caused by faultiness of voice recognition technology in a guide system using a voice recognition technology. <P>SOLUTION: A guide server converts the recorded question voice into text data by performing voice recognizing processing, and displays a voice recognition result correction screen when an unclear part exists (a). Then the question voice is reproduced, and a voice recognition result 127a of an analysis result information 127 is corrected in accordance with correction input of the unclear part "????" by an operator. Then the semantic analysis processing is performed on the text data to determine conditional contents of each retrieval condition, and a semantic analysis result correction screen is displayed when a non-designated retrieval condition exists (b). Then the question voice is reproduced, and the semantic analysis result (127b) of the analysis result information 127 is changed in accordance with options of conditional contents of the selection areas AR2a-AR2d selected by the operator. <P>COPYRIGHT: (C)2006,JPO&NCIPI</p> |
