This paper presents a new statistical language model for speech recognition, based on Generalized LR parsing. The proposed model, the Abstracted Probabilistic GLR (APGLR) model, is an extension of the existing structured language model known as the Probabilistic GLR (PGLR) model. It can predict next words from arbitrarily abstracted categories. The APGLR model is also a generalization of the original PGLR model, because PGLR can be considered to be a special case of APGLRs that predict the next words from the least abstracted grammatical categories, namely the terminal symbols. The selection of the abstraction level is arbitrary; we show several strategies to define the level. The experimental results show that the proposed model performs better than the original PGLR model for speech recognition.
Cite as: Akiba, T., Itou, K. (2001) A structured statistical language model conditioned by arbitrarily abstracted grammatical categories based on GLR parsing. Proc. 7th European Conference on Speech Communication and Technology (Eurospeech 2001), 705-708, doi: 10.21437/Eurospeech.2001-213
@inproceedings{akiba01_eurospeech,
author={Tomoyosi Akiba and Katunobu Itou},
title={{A structured statistical language model conditioned by arbitrarily abstracted grammatical categories based on GLR parsing}},
year=2001,
booktitle={Proc. 7th European Conference on Speech Communication and Technology (Eurospeech 2001)},
pages={705--708},
doi={10.21437/Eurospeech.2001-213}
}