摘要 |
A technique for minimizing the number of users which must dial training digits in a voice processing system. The system uses the first digit normally dialed as part of the user responses to the voice processing application, and measures such digit to determine whether it and future digits can be detected accurately without the use of a training digit. If they can, no training digit is required. If they can't, then a training digit is required, and the user is prompted for the same.
|