Evaluation of feature combination for effective structural disambiguation

So Young Park, Yong Jae Kwak, Joon Ho Lim, Hae-Chang Rim

Research output: Contribution to journalArticle

Abstract

In this paper, we present the useful features of a syntactic constituent for a probabilistic parsing model and analyze the combination of the features in order to disambiguate parse trees effectively. Unlike most of previous works focusing on the features of a single head, the features of a functional head, the features of a content head, and the features of size are utilized in this study. Experimental results show that the combination of different features such as the functional head feature and the size feature is prefered to the combination of similar features such as the functional head feature and the content head feature. Besides, it is remarkable that the function feature is more useful than the combination of the content feature and the size feature.

Original languageEnglish
Pages (from-to)97-101
Number of pages5
JournalLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume2945
Publication statusPublished - 2004 Dec 1

Fingerprint

Syntactics
Head
Evaluation
Parsing
Statistical Models
Experimental Results
Model

ASJC Scopus subject areas

  • Computer Science(all)
  • Biochemistry, Genetics and Molecular Biology(all)
  • Theoretical Computer Science

Cite this

@article{cfc7265e1c6d465ab90ad0106137fa15,
title = "Evaluation of feature combination for effective structural disambiguation",
abstract = "In this paper, we present the useful features of a syntactic constituent for a probabilistic parsing model and analyze the combination of the features in order to disambiguate parse trees effectively. Unlike most of previous works focusing on the features of a single head, the features of a functional head, the features of a content head, and the features of size are utilized in this study. Experimental results show that the combination of different features such as the functional head feature and the size feature is prefered to the combination of similar features such as the functional head feature and the content head feature. Besides, it is remarkable that the function feature is more useful than the combination of the content feature and the size feature.",
author = "Park, {So Young} and Kwak, {Yong Jae} and Lim, {Joon Ho} and Hae-Chang Rim",
year = "2004",
month = "12",
day = "1",
language = "English",
volume = "2945",
pages = "97--101",
journal = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
issn = "0302-9743",
publisher = "Springer Verlag",

}

TY - JOUR

T1 - Evaluation of feature combination for effective structural disambiguation

AU - Park, So Young

AU - Kwak, Yong Jae

AU - Lim, Joon Ho

AU - Rim, Hae-Chang

PY - 2004/12/1

Y1 - 2004/12/1

N2 - In this paper, we present the useful features of a syntactic constituent for a probabilistic parsing model and analyze the combination of the features in order to disambiguate parse trees effectively. Unlike most of previous works focusing on the features of a single head, the features of a functional head, the features of a content head, and the features of size are utilized in this study. Experimental results show that the combination of different features such as the functional head feature and the size feature is prefered to the combination of similar features such as the functional head feature and the content head feature. Besides, it is remarkable that the function feature is more useful than the combination of the content feature and the size feature.

AB - In this paper, we present the useful features of a syntactic constituent for a probabilistic parsing model and analyze the combination of the features in order to disambiguate parse trees effectively. Unlike most of previous works focusing on the features of a single head, the features of a functional head, the features of a content head, and the features of size are utilized in this study. Experimental results show that the combination of different features such as the functional head feature and the size feature is prefered to the combination of similar features such as the functional head feature and the content head feature. Besides, it is remarkable that the function feature is more useful than the combination of the content feature and the size feature.

UR - http://www.scopus.com/inward/record.url?scp=35048823897&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=35048823897&partnerID=8YFLogxK

M3 - Article

VL - 2945

SP - 97

EP - 101

JO - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

JF - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SN - 0302-9743

ER -