To provide a constraint-based speech recognition system and method with high reliability for facilitating speech recognition in a form-filling application employed over a telephone system.

The constraint-based speech recognition system comprises an input signal 36 including both speech input and non-speech input of a type generated by a user via a manually operated device. The system further comprises a constraint module 40 operable to access an information database 44 containing information suitable for speech recognition, and to generated candidate information as a part of the information based on non-speech input and the information database 44. The speech recognition system further comprises a speech recognition module 48 operable to recognize speech based on the speech input and the candidate information. In an exemplary embodiment, the manually operated device is a keypad 26 of a touch-tone telephone 14, and the information database 44 is a lexicon encoded according to classes defined by the keys of the keypad 26 respectively.