Sound signal recognition system and method, and dialog control system and method using it

摘要

A sound signal including either one or both of a voice signal section and a DTMF signal section is inputted to a sound signal matching/recognizing part 200 via a sound signal input part 100. The sound signal is divided into a sound signal section by a sound signal analyzing part 210. A matching part 240 conducts a matching process of a sound signal with reference to both a DTMF signal model and a voice signal model. A recognizing part 260 is provided with a language model 250 including a word dictionary and grammar information, and recognizes a sound signal by using the language model 250 based on the matching result of the matching part 240.