A hybrid convolution tree kernel for semantic role labeling

Wanxiang Che; Min Zhang; Ting Liu; Sheng Li

A hybrid convolution tree kernel for semantic role labeling

Faculty of Computing

Agency for Science, Technology and Research, Singapore
Harbin Institute of Technology

Research output: Contribution to conference › Paper › peer-review

18 Scopus citations

Abstract

A hybrid convolution tree kernel is proposed in this paper to effectively model syntactic structures for semantic role labeling (SRL). The hybrid kernel consists of two individual convolution kernels: a Path kernel, which captures predicate-argument link features, and a Constituent Structure kernel, which captures the syntactic structure features of arguments. Evaluation on the datasets of CoNLL-2005 SRL shared task shows that the novel hybrid convolution tree kernel outperforms the previous tree kernels. We also combine our new hybrid tree kernel based method with the standard rich flat feature based method. The experimental results show that the combinational method can get better performance than each of them individually.

Original language	English
Pages	73-80
Number of pages	8
State	Published - 2006
Event	21st International Conference on Computational Linguistics and 44th Annual Meeting of the Association for Computational Linguistics, COLING/ACL 2006 - Sydney, Australia Duration: 17 Jul 2006 → 18 Jul 2006

Conference

Conference	21st International Conference on Computational Linguistics and 44th Annual Meeting of the Association for Computational Linguistics, COLING/ACL 2006
Country/Territory	Australia
City	Sydney
Period	17/07/06 → 18/07/06

Cite this

@conference{e1a1beabb41b4b5ba21cdf7f1343cf49,

title = "A hybrid convolution tree kernel for semantic role labeling",

abstract = "A hybrid convolution tree kernel is proposed in this paper to effectively model syntactic structures for semantic role labeling (SRL). The hybrid kernel consists of two individual convolution kernels: a Path kernel, which captures predicate-argument link features, and a Constituent Structure kernel, which captures the syntactic structure features of arguments. Evaluation on the datasets of CoNLL-2005 SRL shared task shows that the novel hybrid convolution tree kernel outperforms the previous tree kernels. We also combine our new hybrid tree kernel based method with the standard rich flat feature based method. The experimental results show that the combinational method can get better performance than each of them individually.",

author = "Wanxiang Che and Min Zhang and Ting Liu and Sheng Li",

note = "Publisher Copyright: {\textcopyright} 2006 Association for Computational Linguistics; 21st International Conference on Computational Linguistics and 44th Annual Meeting of the Association for Computational Linguistics, COLING/ACL 2006 ; Conference date: 17-07-2006 Through 18-07-2006",

year = "2006",

language = "英语",

pages = "73--80",

}

A hybrid convolution tree kernel for semantic role labeling. / Che, Wanxiang ; Zhang, Min ; Liu, Ting et al.
2006. 73-80 Paper presented at 21st International Conference on Computational Linguistics and 44th Annual Meeting of the Association for Computational Linguistics, COLING/ACL 2006, Sydney, Australia.

Research output: Contribution to conference › Paper › peer-review

TY - CONF

T1 - A hybrid convolution tree kernel for semantic role labeling

AU - Che, Wanxiang

AU - Zhang, Min

AU - Liu, Ting

AU - Li, Sheng

PY - 2006

Y1 - 2006

N2 - A hybrid convolution tree kernel is proposed in this paper to effectively model syntactic structures for semantic role labeling (SRL). The hybrid kernel consists of two individual convolution kernels: a Path kernel, which captures predicate-argument link features, and a Constituent Structure kernel, which captures the syntactic structure features of arguments. Evaluation on the datasets of CoNLL-2005 SRL shared task shows that the novel hybrid convolution tree kernel outperforms the previous tree kernels. We also combine our new hybrid tree kernel based method with the standard rich flat feature based method. The experimental results show that the combinational method can get better performance than each of them individually.

AB - A hybrid convolution tree kernel is proposed in this paper to effectively model syntactic structures for semantic role labeling (SRL). The hybrid kernel consists of two individual convolution kernels: a Path kernel, which captures predicate-argument link features, and a Constituent Structure kernel, which captures the syntactic structure features of arguments. Evaluation on the datasets of CoNLL-2005 SRL shared task shows that the novel hybrid convolution tree kernel outperforms the previous tree kernels. We also combine our new hybrid tree kernel based method with the standard rich flat feature based method. The experimental results show that the combinational method can get better performance than each of them individually.

UR - https://www.scopus.com/pages/publications/35348959878

M3 - 论文

AN - SCOPUS:35348959878

SP - 73

EP - 80

T2 - 21st International Conference on Computational Linguistics and 44th Annual Meeting of the Association for Computational Linguistics, COLING/ACL 2006

Y2 - 17 July 2006 through 18 July 2006

ER -

A hybrid convolution tree kernel for semantic role labeling

Abstract

Conference

Other files and links

Fingerprint

Cite this