Automatic discovery of part-whole relations

Corina R Girju, Dan Moldovan, Adriana Badulescu

Research output: Contribution to journalArticle

Abstract

An important problem in knowledge discovery from text is the automatic extraction of semantic relations. This paper presents a supervised, semantically intensive, domain independent approach for the automatic detection of part-whole relations in text. First an algorithm is described that identifies lexico-syntactic patterns that encode part-whole relations. A difficulty is that these patterns also encode other semantic relations, and a learning method is necessary to discriminate whether or not a pattern contains a part-whole relation. A large set of training examples have been annotated and fed into a specialized learning system that learns classification rules. The rules are learned through an iterative semantic specialization (ISS) method applied to noun phrase constituents. Classification rules have been generated this way for different patterns such as genitives, noun compounds, and noun phrases containing prepositional phrases to extract part-whole relations from them. The applicability of these rules has been tested on a test corpus obtaining an overall average precision of 80.95% and recall of 75.91 %. The results demonstrate the importance of word sense disambiguation for this task. They also demonstrate that different lexico-syntactic patterns encode different semantic information and should be treated separately in the sense that different clarification rules apply to different patterns.

Original languageEnglish (US)
Pages (from-to)83-135
Number of pages53
JournalComputational Linguistics
Volume32
Issue number1
DOIs
StatePublished - Mar 2006

Fingerprint

Semantics
semantics
Syntactics
learning method
specialization
Data mining
Learning systems
Part-whole Relation
knowledge
learning
Semantic Relations
Noun Phrase
Syntax

ASJC Scopus subject areas

  • Language and Linguistics
  • Linguistics and Language
  • Computer Science Applications
  • Artificial Intelligence

Cite this

Automatic discovery of part-whole relations. / Girju, Corina R; Moldovan, Dan; Badulescu, Adriana.

In: Computational Linguistics, Vol. 32, No. 1, 03.2006, p. 83-135.

Research output: Contribution to journalArticle

Girju, Corina R ; Moldovan, Dan ; Badulescu, Adriana. / Automatic discovery of part-whole relations. In: Computational Linguistics. 2006 ; Vol. 32, No. 1. pp. 83-135.
@article{b55a7b788121460a9c5d6f0b86785e98,
title = "Automatic discovery of part-whole relations",
abstract = "An important problem in knowledge discovery from text is the automatic extraction of semantic relations. This paper presents a supervised, semantically intensive, domain independent approach for the automatic detection of part-whole relations in text. First an algorithm is described that identifies lexico-syntactic patterns that encode part-whole relations. A difficulty is that these patterns also encode other semantic relations, and a learning method is necessary to discriminate whether or not a pattern contains a part-whole relation. A large set of training examples have been annotated and fed into a specialized learning system that learns classification rules. The rules are learned through an iterative semantic specialization (ISS) method applied to noun phrase constituents. Classification rules have been generated this way for different patterns such as genitives, noun compounds, and noun phrases containing prepositional phrases to extract part-whole relations from them. The applicability of these rules has been tested on a test corpus obtaining an overall average precision of 80.95{\%} and recall of 75.91 {\%}. The results demonstrate the importance of word sense disambiguation for this task. They also demonstrate that different lexico-syntactic patterns encode different semantic information and should be treated separately in the sense that different clarification rules apply to different patterns.",
author = "Girju, {Corina R} and Dan Moldovan and Adriana Badulescu",
year = "2006",
month = "3",
doi = "10.1162/coli.2006.32.1.83",
language = "English (US)",
volume = "32",
pages = "83--135",
journal = "Computational Linguistics",
issn = "0891-2017",
publisher = "MIT Press Journals",
number = "1",

}

TY - JOUR

T1 - Automatic discovery of part-whole relations

AU - Girju, Corina R

AU - Moldovan, Dan

AU - Badulescu, Adriana

PY - 2006/3

Y1 - 2006/3

N2 - An important problem in knowledge discovery from text is the automatic extraction of semantic relations. This paper presents a supervised, semantically intensive, domain independent approach for the automatic detection of part-whole relations in text. First an algorithm is described that identifies lexico-syntactic patterns that encode part-whole relations. A difficulty is that these patterns also encode other semantic relations, and a learning method is necessary to discriminate whether or not a pattern contains a part-whole relation. A large set of training examples have been annotated and fed into a specialized learning system that learns classification rules. The rules are learned through an iterative semantic specialization (ISS) method applied to noun phrase constituents. Classification rules have been generated this way for different patterns such as genitives, noun compounds, and noun phrases containing prepositional phrases to extract part-whole relations from them. The applicability of these rules has been tested on a test corpus obtaining an overall average precision of 80.95% and recall of 75.91 %. The results demonstrate the importance of word sense disambiguation for this task. They also demonstrate that different lexico-syntactic patterns encode different semantic information and should be treated separately in the sense that different clarification rules apply to different patterns.

AB - An important problem in knowledge discovery from text is the automatic extraction of semantic relations. This paper presents a supervised, semantically intensive, domain independent approach for the automatic detection of part-whole relations in text. First an algorithm is described that identifies lexico-syntactic patterns that encode part-whole relations. A difficulty is that these patterns also encode other semantic relations, and a learning method is necessary to discriminate whether or not a pattern contains a part-whole relation. A large set of training examples have been annotated and fed into a specialized learning system that learns classification rules. The rules are learned through an iterative semantic specialization (ISS) method applied to noun phrase constituents. Classification rules have been generated this way for different patterns such as genitives, noun compounds, and noun phrases containing prepositional phrases to extract part-whole relations from them. The applicability of these rules has been tested on a test corpus obtaining an overall average precision of 80.95% and recall of 75.91 %. The results demonstrate the importance of word sense disambiguation for this task. They also demonstrate that different lexico-syntactic patterns encode different semantic information and should be treated separately in the sense that different clarification rules apply to different patterns.

UR - http://www.scopus.com/inward/record.url?scp=33646811217&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33646811217&partnerID=8YFLogxK

U2 - 10.1162/coli.2006.32.1.83

DO - 10.1162/coli.2006.32.1.83

M3 - Article

AN - SCOPUS:33646811217

VL - 32

SP - 83

EP - 135

JO - Computational Linguistics

JF - Computational Linguistics

SN - 0891-2017

IS - 1

ER -