A high accuracy method for semi-supervised information extraction

Stephen Tratz; Antonio Sanfilippo

doi:10.3115/1614108.1614151

A high accuracy method for semi-supervised information extraction

Stephen Tratz, Antonio Sanfilippo

Pacific Northwest National Laboratory

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

5 Citations (Scopus)

Abstract

Customization to specific domains of discourse and/or user requirements is one of the greatest challenges for today's Information Extraction (IE) systems. While demonstrably effective, both rule-based and supervised machine learning approaches to IE customization pose too high a burden on the user. Semisupervised learning approaches may in principle offer a more resource effective solution but are still insufficiently accurate to grant realistic application. We demonstrate that this limitation can be overcome by integrating fully-supervised learning techniques within a semisupervised IE approach, without increasing resource requirements.

Original language	English
Pages	169-172
Number of pages	4
DOIs	https://doi.org/10.3115/1614108.1614151
Publication status	Published - 2007
Externally published	Yes
Event	2007 Human Language Technology Conference of the North American Chapter of the Association of Computational Linguistics, NAACL-HLT 2007 - Rochester, United States Duration: 22 Apr 2007 → 27 Apr 2007

Conference

Conference	2007 Human Language Technology Conference of the North American Chapter of the Association of Computational Linguistics, NAACL-HLT 2007
Country/Territory	United States
City	Rochester
Period	22/04/07 → 27/04/07

Access to Document

10.3115/1614108.1614151

Cite this

@conference{5894538aa5204a389f37f635fc18d0cc,

title = "A high accuracy method for semi-supervised information extraction",

abstract = "Customization to specific domains of discourse and/or user requirements is one of the greatest challenges for today's Information Extraction (IE) systems. While demonstrably effective, both rule-based and supervised machine learning approaches to IE customization pose too high a burden on the user. Semisupervised learning approaches may in principle offer a more resource effective solution but are still insufficiently accurate to grant realistic application. We demonstrate that this limitation can be overcome by integrating fully-supervised learning techniques within a semisupervised IE approach, without increasing resource requirements.",

author = "Stephen Tratz and Antonio Sanfilippo",

note = "Publisher Copyright: {\textcopyright} 2007 Association for Computational Linguistics.; 2007 Human Language Technology Conference of the North American Chapter of the Association of Computational Linguistics, NAACL-HLT 2007 ; Conference date: 22-04-2007 Through 27-04-2007",

year = "2007",

doi = "10.3115/1614108.1614151",

language = "English",

pages = "169--172",

}

A high accuracy method for semi-supervised information extraction. / Tratz, Stephen; Sanfilippo, Antonio.
2007. 169-172 Paper presented at 2007 Human Language Technology Conference of the North American Chapter of the Association of Computational Linguistics, NAACL-HLT 2007, Rochester, United States.

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

TY - CONF

T1 - A high accuracy method for semi-supervised information extraction

AU - Tratz, Stephen

AU - Sanfilippo, Antonio

PY - 2007

Y1 - 2007

N2 - Customization to specific domains of discourse and/or user requirements is one of the greatest challenges for today's Information Extraction (IE) systems. While demonstrably effective, both rule-based and supervised machine learning approaches to IE customization pose too high a burden on the user. Semisupervised learning approaches may in principle offer a more resource effective solution but are still insufficiently accurate to grant realistic application. We demonstrate that this limitation can be overcome by integrating fully-supervised learning techniques within a semisupervised IE approach, without increasing resource requirements.

AB - Customization to specific domains of discourse and/or user requirements is one of the greatest challenges for today's Information Extraction (IE) systems. While demonstrably effective, both rule-based and supervised machine learning approaches to IE customization pose too high a burden on the user. Semisupervised learning approaches may in principle offer a more resource effective solution but are still insufficiently accurate to grant realistic application. We demonstrate that this limitation can be overcome by integrating fully-supervised learning techniques within a semisupervised IE approach, without increasing resource requirements.

UR - http://www.scopus.com/inward/record.url?scp=84864676115&partnerID=8YFLogxK

U2 - 10.3115/1614108.1614151

DO - 10.3115/1614108.1614151

M3 - Paper

AN - SCOPUS:84864676115

SP - 169

EP - 172

T2 - 2007 Human Language Technology Conference of the North American Chapter of the Association of Computational Linguistics, NAACL-HLT 2007

Y2 - 22 April 2007 through 27 April 2007

ER -

A high accuracy method for semi-supervised information extraction

Abstract

Conference

Access to Document

Other files and links

Fingerprint

Cite this