Belief Networks for A Syntactic and Semantic Analyiss of Spoken Utterances for Speech Understanding

摘要

In this paper we present a new approach towards speech understanding that merges semantic and intention decoding to oen component. The algorithm is supposed to evalaute a speech recognizer's utterance hypotheses regarding a) syntactical and semantical relations between words and phrases and b) potential intentions of hte user. The mathematical fundament for this evalaution is probability theory. We make use of belief networks to handle the anlaysis of an utterance hypothesis as a process of reasoning with uncertain and incomplete information. The algorithm in general can be characterized as phrase spotting.