| 摘要 |
<p><P>PROBLEM TO BE SOLVED: To dispense with troublesome operation input to achieve quick and reliable refinement, when retrieving a content desired by a user from a content group. <P>SOLUTION: An agent 1 has a voice recognition part 3, a voice composition part 4 with a control circuit 2, can access a server 13 connected to the Internet through a communication device 7, and retrieves databases 14a-14c registering the content group. The databases 14a-14c are incidentally registered with meta information having an answer to a question item for retrieving the content together with information about the content. The agent 1 outputs the question item from a speaker 11 by voice composition, performs voice recognition when the user answers the number of the answer by voice, and specifies the answer. The retrieval processing is executed on the basis of the obtained answer, and the refinement is performed by repeating it. <P>COPYRIGHT: (C)2008,JPO&INPIT</p> |
