In this paper, a special text processing method is presented, which integrates external descriptor rule and model matching. The letter-to-sound rules for a given kind of special mark can be described by some describable features which can be extracted from this kind of special mark, and then a corresponding model can be built and then the model parameter can be stored into a external form. Since the features can be extracted and described conveniently, a new model can be built easily for improving the ability to analyze special text.
Cite as: Wu, X., Wang, R., Hu, G. (2000) Special text processing based external descriptor rule. Proc. 6th International Conference on Spoken Language Processing (ICSLP 2000), vol. 1, 689-692
@inproceedings{wu00b_icslp,
author={Xiaoru Wu and Renhua Wang and Guoping Hu},
title={{Special text processing based external descriptor rule}},
year=2000,
booktitle={Proc. 6th International Conference on Spoken Language Processing (ICSLP 2000)},
pages={vol. 1, 689-692}
}