摘要 |
PURPOSE: A speech recognition system applying a client/server model and a method for providing a speech recognition service are provided to individually maintain and manage a speech recognition system by independently maintaining a speech recognition function. CONSTITUTION: A speech recognition service client controls a telephone network interface and a scenario and security database. A recognition server processes a speech recognition function, and transmits a recognition result to a client using a TCP/IP. A telephone network interface device transmits speech information generated from the result to a user through a telephone network. A TCP/IP client device transmits a feature value extracted to the recognition server using the TCP/IP, and transmits the result to a scenario management device. The scenario management device collects information from a content database according to the result, and generates speech information. The scenario management device is included in the speech recognition service client.
|