Probabilistic LR Parsing for Speech Recognition

Abstract

An LR parser for probabilistic context-free grammars is described. Each of the standard versions of parser generator (SLR, canonical and LALR) may be applied. A graph-structured stack permits action conflicts and allows the parser to be used with uncertain input, typical of speech recognition applications. The sentence uncertainty is measured using entropy and is significantly lower for the grammar than for a first-order Markov model.