ZČU
Publications
Detail of publication
Citation
: Fitting class-based language models into weighted finite-state transducer framework . Eurospeech, vol. 1, p. 1873-1876, ISCA, Geneva, 2003.
Abstract
In our paper we propose a general way of incorporating class-based language models with many-to-many word-to-class mapping into the finite-state transducer (FST) framework. Since class-based models alone usually do not improve the recognition accuracy, we also present a method for an efficient language model combination. An example of a word-to-class mapping based on morphological tags is also given. Several word-based and tag-based language models are tested in the task of transcribing Czech broadcast news. Results show that class-based models help to achieve a moderate improvement in recognition accuracy.
Detail of publication
| Title: | Fitting class-based language models into weighted finite-state transducer framework |
|---|---|
| Author: | Ircing, P. ; Psutka, J. |
| Language: | English |
| Date of publication: | 1 Sep 2003 |
| Year: | 2003 |
| Type of publication: | Papers in journals |
| Title of journal or book: | Eurospeech |
| Číslo vydání: | 1 |
| Page: | 1873 - 1876 |
| ISBN: | 1018-4074 |
| ISSN: | 1018-4074 |
| Publisher: | ISCA |
| Address: | Geneva |
| Date: | 1 Sep 2003 - 4 Sep 2003 |
Keywords
language modeling, finite-state automata, speech recognition
BibTeX
@ARTICLE{IrcingP_2003_Fittingclass-based_1,
author = {Ircing, P. and Psutka, J.},
title = {Fitting class-based language models into weighted finite-state transducer framework},
year = {2003},
publisher = {ISCA},
journal = {Eurospeech},
address = {Geneva},
volume = {1},
pages = {1873-1876},
ISBN = {1018-4074},
ISSN = {1018-4074},
url = {http://www.kky.zcu.cz/en/publications/IrcingP_2003_Fittingclass-based_1},
}