Joint Parsing and Semantic Role Labeling

Charles Sutton; Andrew McCallum

Joint Parsing and Semantic Role Labeling

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

A striking feature of human syntactic processing is that it is context-dependent, that is, it seems to take into account semantic information from the discourse context and world knowledge. In this paper, we attempt to use this insight to bridge the gap between SRL results from gold parses and from automatically-generated parses. To do this, we jointly perform parsing and semantic role labeling, using a probabilistic SRL system to rerank the results of a probabilistic parser. Our current results are negative, because a locally-trained SRL model can return inaccurate probability estimates.

Original language	English
Title of host publication	Proceedings of the Ninth Conference on Computational Natural Language Learning (CoNLL-2005)
Place of Publication	Ann Arbor, Michigan
Publisher	Association for Computational Linguistics
Pages	225-228
Number of pages	4
Publication status	Published - 1 Jun 2005

Access to Document

W05-0636Final published version, 84.9 KB

http://www.aclweb.org/anthology/W/W05/W05-0636

Fingerprint Dive into the research topics of 'Joint Parsing and Semantic Role Labeling'. Together they form a unique fingerprint.

Cite this

@inproceedings{661da25fee814c8a91edeef461033727,

title = "Joint Parsing and Semantic Role Labeling",

abstract = "A striking feature of human syntactic processing is that it is context-dependent, that is, it seems to take into account semantic information from the discourse context and world knowledge. In this paper, we attempt to use this insight to bridge the gap between SRL results from gold parses and from automatically-generated parses. To do this, we jointly perform parsing and semantic role labeling, using a probabilistic SRL system to rerank the results of a probabilistic parser. Our current results are negative, because a locally-trained SRL model can return inaccurate probability estimates.",

author = "Charles Sutton and Andrew McCallum",

year = "2005",

month = jun,

day = "1",

language = "English",

pages = "225--228",

booktitle = "Proceedings of the Ninth Conference on Computational Natural Language Learning (CoNLL-2005)",

publisher = "Association for Computational Linguistics",

}

TY - GEN

T1 - Joint Parsing and Semantic Role Labeling

AU - Sutton, Charles

AU - McCallum, Andrew

PY - 2005/6/1

Y1 - 2005/6/1

N2 - A striking feature of human syntactic processing is that it is context-dependent, that is, it seems to take into account semantic information from the discourse context and world knowledge. In this paper, we attempt to use this insight to bridge the gap between SRL results from gold parses and from automatically-generated parses. To do this, we jointly perform parsing and semantic role labeling, using a probabilistic SRL system to rerank the results of a probabilistic parser. Our current results are negative, because a locally-trained SRL model can return inaccurate probability estimates.

AB - A striking feature of human syntactic processing is that it is context-dependent, that is, it seems to take into account semantic information from the discourse context and world knowledge. In this paper, we attempt to use this insight to bridge the gap between SRL results from gold parses and from automatically-generated parses. To do this, we jointly perform parsing and semantic role labeling, using a probabilistic SRL system to rerank the results of a probabilistic parser. Our current results are negative, because a locally-trained SRL model can return inaccurate probability estimates.

M3 - Conference contribution

SP - 225

EP - 228

BT - Proceedings of the Ninth Conference on Computational Natural Language Learning (CoNLL-2005)

PB - Association for Computational Linguistics

CY - Ann Arbor, Michigan

ER -