ML-Tuned Constraint Grammars

Eckhard Bick

ML-Tuned Constraint Grammars

Eckhard Bick

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

Abstract

In this paper we present a new method for machine learning-based optimization of linguist-written Constraint Grammars. The effect of rule ordering/sorting, grammarsectioning and systematic rule changes is discussed and quantitatively evaluated. The F-score improvement was 0.41 percentage points for a mature (Danish) tagging grammar, and 1.36 percentage points for a half-size grammar, translating into a 7-15% error reduction relative to the performance of the untuned grammars.

Original language	English
Title of host publication	Proceedings of the 27th Pacific Asia Conference on Language, Information and Computation
Place of Publication	Taipei
Publisher	Department of English, National Chengchi University
Publication date	2013
Pages	440-449
ISBN (Electronic)	978-986-03-8567-0
Publication status	Published - 2013
Event	PACLIC 2013: The 27th Pacific Asia Conference on Language, Information, and Computation - National Chengchi University, No. 64, Sec. 2, Zhinan Rd., Wenshan,Taipei, Taiwan, Taipei, Taiwan Duration: 21. Nov 2013 → 24. Nov 2013 Conference number: 27

Conference

Conference	PACLIC 2013
Number	27
Location	National Chengchi University, No. 64, Sec. 2, Zhinan Rd., Wenshan,Taipei, Taiwan
Country/Territory	Taiwan
City	Taipei
Period	21/11/2013 → 24/11/2013

SDU Link

Check access to the material in SDU Library's search engine

PACLIC 2013
Bick, E. (Speaker)
Nov 2014 → …
Activity: Talks and presentations › Conference presentations

Cite this

@inproceedings{bc191a4d17cc4d699d22dd80789914f6,

title = "ML-Tuned Constraint Grammars",

abstract = "In this paper we present a new method for machine learning-based optimization of linguist-written Constraint Grammars. The effect of rule ordering/sorting, grammarsectioning and systematic rule changes is discussed and quantitatively evaluated. The F-score improvement was 0.41 percentage points for a mature (Danish) tagging grammar, and 1.36 percentage points for a half-size grammar, translating into a 7-15\% error reduction relative to the performance of the untuned grammars.",

keywords = "Constraint Grammar, Machine Learning, Parsing, Tagging",

author = "Eckhard Bick",

year = "2013",

language = "English",

pages = "440--449",

booktitle = "Proceedings of the 27th Pacific Asia Conference on Language, Information and Computation",

publisher = "Department of English, National Chengchi University",

note = "PACLIC 2013 ; Conference date: 21-11-2013 Through 24-11-2013",

}

TY - GEN

T1 - ML-Tuned Constraint Grammars

AU - Bick, Eckhard

N1 - Conference code: 27

PY - 2013

Y1 - 2013

N2 - In this paper we present a new method for machine learning-based optimization of linguist-written Constraint Grammars. The effect of rule ordering/sorting, grammarsectioning and systematic rule changes is discussed and quantitatively evaluated. The F-score improvement was 0.41 percentage points for a mature (Danish) tagging grammar, and 1.36 percentage points for a half-size grammar, translating into a 7-15% error reduction relative to the performance of the untuned grammars.

AB - In this paper we present a new method for machine learning-based optimization of linguist-written Constraint Grammars. The effect of rule ordering/sorting, grammarsectioning and systematic rule changes is discussed and quantitatively evaluated. The F-score improvement was 0.41 percentage points for a mature (Danish) tagging grammar, and 1.36 percentage points for a half-size grammar, translating into a 7-15% error reduction relative to the performance of the untuned grammars.

KW - Constraint Grammar

KW - Machine Learning

KW - Parsing

KW - Tagging

M3 - Article in proceedings

SP - 440

EP - 449

BT - Proceedings of the 27th Pacific Asia Conference on Language, Information and Computation

PB - Department of English, National Chengchi University

CY - Taipei

T2 - PACLIC 2013

Y2 - 21 November 2013 through 24 November 2013

ER -

ML-Tuned Constraint Grammars

Abstract

Conference

SDU Link

Fingerprint

Related activities

PACLIC 2013

Cite this