Dependency parse reranking with rich subtree features

Mo Shen, Daisuke Kawahara, Sadao Kurohashi

研究成果: ジャーナルへの寄稿記事

3 引用 (Scopus)

抄録

In pursuing machine understanding of human language, highly accurate syntactic analysis is a crucial step. In this work, we focus on dependency grammar, which models syntax by encoding transparent predicate-argument structures. Recent advances in dependency parsing have shown that employing higherorder subtree structures in graph-based parsers can substantially improve the parsing accuracy. However, the inefficiency of this approach increases with the order of the subtrees. This work explores a new reranking approach for dependency parsing that can utilize complex subtree representations by applying efficient subtree selection methods. We demonstrate the effectiveness of the approach in experiments conducted on the Penn Treebank and the Chinese Treebank. Our system achieves the best performance among known supervised systems evaluated on these datasets, improving the baseline accuracy from 91.88% to 93.42% for English, and from 87.39% to 89.25% for Chinese.

元の言語英語
記事番号2327295
ページ(範囲)1208-1218
ページ数11
ジャーナルIEEE Transactions on Audio, Speech and Language Processing
22
発行部数7
DOI
出版物ステータス出版済み - 7 1 2014

Fingerprint

Syntactics
syntax
grammars
Experiments
coding

All Science Journal Classification (ASJC) codes

  • Acoustics and Ultrasonics
  • Electrical and Electronic Engineering

これを引用

Dependency parse reranking with rich subtree features. / Shen, Mo; Kawahara, Daisuke; Kurohashi, Sadao.

:: IEEE Transactions on Audio, Speech and Language Processing, 巻 22, 番号 7, 2327295, 01.07.2014, p. 1208-1218.

研究成果: ジャーナルへの寄稿記事

Shen, Mo ; Kawahara, Daisuke ; Kurohashi, Sadao. / Dependency parse reranking with rich subtree features. :: IEEE Transactions on Audio, Speech and Language Processing. 2014 ; 巻 22, 番号 7. pp. 1208-1218.
@article{f7f012c3a6a44d24813de571ac28aff6,
title = "Dependency parse reranking with rich subtree features",
abstract = "In pursuing machine understanding of human language, highly accurate syntactic analysis is a crucial step. In this work, we focus on dependency grammar, which models syntax by encoding transparent predicate-argument structures. Recent advances in dependency parsing have shown that employing higherorder subtree structures in graph-based parsers can substantially improve the parsing accuracy. However, the inefficiency of this approach increases with the order of the subtrees. This work explores a new reranking approach for dependency parsing that can utilize complex subtree representations by applying efficient subtree selection methods. We demonstrate the effectiveness of the approach in experiments conducted on the Penn Treebank and the Chinese Treebank. Our system achieves the best performance among known supervised systems evaluated on these datasets, improving the baseline accuracy from 91.88{\%} to 93.42{\%} for English, and from 87.39{\%} to 89.25{\%} for Chinese.",
author = "Mo Shen and Daisuke Kawahara and Sadao Kurohashi",
year = "2014",
month = "7",
day = "1",
doi = "10.1109/TASLP.2014.2327295",
language = "English",
volume = "22",
pages = "1208--1218",
journal = "IEEE Transactions on Speech and Audio Processing",
issn = "1558-7916",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "7",

}

TY - JOUR

T1 - Dependency parse reranking with rich subtree features

AU - Shen, Mo

AU - Kawahara, Daisuke

AU - Kurohashi, Sadao

PY - 2014/7/1

Y1 - 2014/7/1

N2 - In pursuing machine understanding of human language, highly accurate syntactic analysis is a crucial step. In this work, we focus on dependency grammar, which models syntax by encoding transparent predicate-argument structures. Recent advances in dependency parsing have shown that employing higherorder subtree structures in graph-based parsers can substantially improve the parsing accuracy. However, the inefficiency of this approach increases with the order of the subtrees. This work explores a new reranking approach for dependency parsing that can utilize complex subtree representations by applying efficient subtree selection methods. We demonstrate the effectiveness of the approach in experiments conducted on the Penn Treebank and the Chinese Treebank. Our system achieves the best performance among known supervised systems evaluated on these datasets, improving the baseline accuracy from 91.88% to 93.42% for English, and from 87.39% to 89.25% for Chinese.

AB - In pursuing machine understanding of human language, highly accurate syntactic analysis is a crucial step. In this work, we focus on dependency grammar, which models syntax by encoding transparent predicate-argument structures. Recent advances in dependency parsing have shown that employing higherorder subtree structures in graph-based parsers can substantially improve the parsing accuracy. However, the inefficiency of this approach increases with the order of the subtrees. This work explores a new reranking approach for dependency parsing that can utilize complex subtree representations by applying efficient subtree selection methods. We demonstrate the effectiveness of the approach in experiments conducted on the Penn Treebank and the Chinese Treebank. Our system achieves the best performance among known supervised systems evaluated on these datasets, improving the baseline accuracy from 91.88% to 93.42% for English, and from 87.39% to 89.25% for Chinese.

UR - http://www.scopus.com/inward/record.url?scp=84911390605&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84911390605&partnerID=8YFLogxK

U2 - 10.1109/TASLP.2014.2327295

DO - 10.1109/TASLP.2014.2327295

M3 - Article

AN - SCOPUS:84911390605

VL - 22

SP - 1208

EP - 1218

JO - IEEE Transactions on Speech and Audio Processing

JF - IEEE Transactions on Speech and Audio Processing

SN - 1558-7916

IS - 7

M1 - 2327295

ER -