Stochastic language models for speech recognition have traditionally been designed and evaluated in order to optimize word accuracy. In this work, we present a novel framework for training stochastic language models by optimizing two different criteria appropriate for speech recognition and language understanding. First, the language entropy and "salience" measure are used for learning the "relevant" spoken language features (phrases). Secondly, a novel algorithm for training stochastic finite state machines is presented which incorporates the acquired phrase structure into a single stochastic language model. Thirdly, we show the benefit of our novel framework with an end-to-end evaluation of a large vocabulary spoken language system for call routing.
Cite as: Riccardi, G., Gorin, A.L. (1998) Stochastic language models for speech recognition and understanding. Proc. 5th International Conference on Spoken Language Processing (ICSLP 1998), paper 0111, doi: 10.21437/ICSLP.1998-502
@inproceedings{riccardi98b_icslp,
author={Giuseppe Riccardi and Allen L. Gorin},
title={{Stochastic language models for speech recognition and understanding}},
year=1998,
booktitle={Proc. 5th International Conference on Spoken Language Processing (ICSLP 1998)},
pages={paper 0111},
doi={10.21437/ICSLP.1998-502}
}