Feature-rich part-of-speech tagging with a cyclic dependency network
K. Toutanova, D. Klein, C. Manning, and Y. Singer. NAACL '03: Proceedings of the 2003 Conference of the North American Chapter of the Association for Computational Linguistics on Human Language Technology, page 173--180. Morristown, NJ, USA, Association for Computational Linguistics, (2003)
DOI: http://dx.doi.org/10.3115/1073445.1073478
Abstract
We present a new part-of-speech tagger that demonstrates the following ideas: (i) explicit use of both preceding and following tag contexts via a dependency network representation, (ii) broad use of lexical features, including jointly conditioning on multiple consecutive words, (iii) effective use of priors in conditional loglinear models, and (iv) fine-grained modeling of unknown word features. Using these ideas together, the resulting tagger gives a 97.24% accuracy on the Penn Treebank WSJ, an error reduction of 4.4% on the best previous single automatically learned tagging result.
Description
Feature-rich part-of-speech tagging with a cyclic dependency network
NAACL '03: Proceedings of the 2003 Conference of the North American Chapter of the Association for Computational Linguistics on Human Language Technology
%0 Conference Paper
%1 Toutanova03posTagging
%A Toutanova, Kristina
%A Klein, Dan
%A Manning, Christopher D.
%A Singer, Yoram
%B NAACL '03: Proceedings of the 2003 Conference of the North American Chapter of the Association for Computational Linguistics on Human Language Technology
%C Morristown, NJ, USA
%D 2003
%I Association for Computational Linguistics
%K 03 Toutanova pos stanford tagging
%P 173--180
%R http://dx.doi.org/10.3115/1073445.1073478
%T Feature-rich part-of-speech tagging with a cyclic dependency network
%U http://portal.acm.org/citation.cfm?id=1073445.1073478
%X We present a new part-of-speech tagger that demonstrates the following ideas: (i) explicit use of both preceding and following tag contexts via a dependency network representation, (ii) broad use of lexical features, including jointly conditioning on multiple consecutive words, (iii) effective use of priors in conditional loglinear models, and (iv) fine-grained modeling of unknown word features. Using these ideas together, the resulting tagger gives a 97.24% accuracy on the Penn Treebank WSJ, an error reduction of 4.4% on the best previous single automatically learned tagging result.
@inproceedings{Toutanova03posTagging,
abstract = {We present a new part-of-speech tagger that demonstrates the following ideas: (i) explicit use of both preceding and following tag contexts via a dependency network representation, (ii) broad use of lexical features, including jointly conditioning on multiple consecutive words, (iii) effective use of priors in conditional loglinear models, and (iv) fine-grained modeling of unknown word features. Using these ideas together, the resulting tagger gives a 97.24% accuracy on the Penn Treebank WSJ, an error reduction of 4.4% on the best previous single automatically learned tagging result.},
added-at = {2010-03-10T18:39:02.000+0100},
address = {Morristown, NJ, USA},
author = {Toutanova, Kristina and Klein, Dan and Manning, Christopher D. and Singer, Yoram},
biburl = {https://www.bibsonomy.org/bibtex/25b0463528f0366670e9c003ba277df05/lee_peck},
booktitle = {NAACL '03: Proceedings of the 2003 Conference of the North American Chapter of the Association for Computational Linguistics on Human Language Technology},
description = {Feature-rich part-of-speech tagging with a cyclic dependency network},
doi = {http://dx.doi.org/10.3115/1073445.1073478},
interhash = {10c3b8e87b2447af2ae27bdee2cdae7f},
intrahash = {5b0463528f0366670e9c003ba277df05},
keywords = {03 Toutanova pos stanford tagging},
location = {Edmonton, Canada},
pages = {173--180},
publisher = {Association for Computational Linguistics},
timestamp = {2010-03-10T18:39:02.000+0100},
title = {Feature-rich part-of-speech tagging with a cyclic dependency network},
url = {http://portal.acm.org/citation.cfm?id=1073445.1073478},
year = 2003
}