A reranking approach for dependency parsing with variable-sized subtree features

Mo Shen, Daisuke Kawahara, Sadao Kurohashi

Research output: Chapter in Book/Report/Conference proceedingConference contribution

5 Citations (Scopus)

Abstract

Employing higher-order subtree structures in graph-based dependency parsing has shown substantial improvement over the accuracy, however suffers from the inefficiency increasing with the order of subtrees. We present a new reranking approach for dependency parsing that can utilize complex subtree representation by applying efficient subtree selection heuristics. We demonstrate the effectiveness of the approach in experiments conducted on the Penn Treebank and the Chinese Treebank. Our system improves the baseline accuracy from 91.88% to 93.37% for English, and in the case of Chinese from 87.39% to 89.16%.

Original languageEnglish
Title of host publicationProceedings of the 26th Pacific Asia Conference on Language, Information and Computation, PACLIC 2012
Pages308-317
Number of pages10
Publication statusPublished - Dec 1 2012
Event26th Pacific Asia Conference on Language, Information and Computation, PACLIC 2012 - Bali, Indonesia
Duration: Nov 7 2012Nov 7 2012

Other

Other26th Pacific Asia Conference on Language, Information and Computation, PACLIC 2012
CountryIndonesia
CityBali
Period11/7/1211/7/12

Fingerprint

Experiments

All Science Journal Classification (ASJC) codes

  • Information Systems
  • Software

Cite this

Shen, M., Kawahara, D., & Kurohashi, S. (2012). A reranking approach for dependency parsing with variable-sized subtree features. In Proceedings of the 26th Pacific Asia Conference on Language, Information and Computation, PACLIC 2012 (pp. 308-317)

A reranking approach for dependency parsing with variable-sized subtree features. / Shen, Mo; Kawahara, Daisuke; Kurohashi, Sadao.

Proceedings of the 26th Pacific Asia Conference on Language, Information and Computation, PACLIC 2012. 2012. p. 308-317.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Shen, M, Kawahara, D & Kurohashi, S 2012, A reranking approach for dependency parsing with variable-sized subtree features. in Proceedings of the 26th Pacific Asia Conference on Language, Information and Computation, PACLIC 2012. pp. 308-317, 26th Pacific Asia Conference on Language, Information and Computation, PACLIC 2012, Bali, Indonesia, 11/7/12.
Shen M, Kawahara D, Kurohashi S. A reranking approach for dependency parsing with variable-sized subtree features. In Proceedings of the 26th Pacific Asia Conference on Language, Information and Computation, PACLIC 2012. 2012. p. 308-317
Shen, Mo ; Kawahara, Daisuke ; Kurohashi, Sadao. / A reranking approach for dependency parsing with variable-sized subtree features. Proceedings of the 26th Pacific Asia Conference on Language, Information and Computation, PACLIC 2012. 2012. pp. 308-317
@inproceedings{cebad8848617434fbc5e0e87d3f95ebd,
title = "A reranking approach for dependency parsing with variable-sized subtree features",
abstract = "Employing higher-order subtree structures in graph-based dependency parsing has shown substantial improvement over the accuracy, however suffers from the inefficiency increasing with the order of subtrees. We present a new reranking approach for dependency parsing that can utilize complex subtree representation by applying efficient subtree selection heuristics. We demonstrate the effectiveness of the approach in experiments conducted on the Penn Treebank and the Chinese Treebank. Our system improves the baseline accuracy from 91.88{\%} to 93.37{\%} for English, and in the case of Chinese from 87.39{\%} to 89.16{\%}.",
author = "Mo Shen and Daisuke Kawahara and Sadao Kurohashi",
year = "2012",
month = "12",
day = "1",
language = "English",
isbn = "9789791421171",
pages = "308--317",
booktitle = "Proceedings of the 26th Pacific Asia Conference on Language, Information and Computation, PACLIC 2012",

}

TY - GEN

T1 - A reranking approach for dependency parsing with variable-sized subtree features

AU - Shen, Mo

AU - Kawahara, Daisuke

AU - Kurohashi, Sadao

PY - 2012/12/1

Y1 - 2012/12/1

N2 - Employing higher-order subtree structures in graph-based dependency parsing has shown substantial improvement over the accuracy, however suffers from the inefficiency increasing with the order of subtrees. We present a new reranking approach for dependency parsing that can utilize complex subtree representation by applying efficient subtree selection heuristics. We demonstrate the effectiveness of the approach in experiments conducted on the Penn Treebank and the Chinese Treebank. Our system improves the baseline accuracy from 91.88% to 93.37% for English, and in the case of Chinese from 87.39% to 89.16%.

AB - Employing higher-order subtree structures in graph-based dependency parsing has shown substantial improvement over the accuracy, however suffers from the inefficiency increasing with the order of subtrees. We present a new reranking approach for dependency parsing that can utilize complex subtree representation by applying efficient subtree selection heuristics. We demonstrate the effectiveness of the approach in experiments conducted on the Penn Treebank and the Chinese Treebank. Our system improves the baseline accuracy from 91.88% to 93.37% for English, and in the case of Chinese from 87.39% to 89.16%.

UR - http://www.scopus.com/inward/record.url?scp=84883420449&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84883420449&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84883420449

SN - 9789791421171

SP - 308

EP - 317

BT - Proceedings of the 26th Pacific Asia Conference on Language, Information and Computation, PACLIC 2012

ER -