Optimisation of corpus-derived probabilistic grammars

Research output: Chapter in Book/Conference proceeding with ISSN or ISBNConference contribution with ISSN or ISBNResearchpeer-review

Abstract

This paper examines the usefulness of corpus-derived probabilistic grammars as a basis for the automatic construction of grammars optimised for a given parsing task. Initially, a probabilistic context-free grammar (PCFG) is derived by a straightforward derivation technique from the Wall Street Journal (WSJ) Corpus, and a baseline is established by testing the resulting grammar on four different parsing tasks. In the first optimisation step, different kinds of local structural context (LSC) are incorporated into the basic PCFG. Improved parsing results demonstrate the usefulness of the added structural context information. In the second optimisation step, LSC-PCFGs are optimised in terms of grammar size and performance for a given parsing task. Tests show that significant improvements can be achieved by the method proposed. The structure of this paper is as follows. Section 2 discusses the practica
Original languageEnglish
Title of host publicationCorpus Linguistics 2001
Place of PublicationLancaster, UK
Pages46-57
Number of pages12
DOIs
Publication statusPublished - 1 Mar 2001
EventCorpus Linguistics 2001 - Lancaster University, UK
Duration: 1 Mar 2001 → …

Conference

ConferenceCorpus Linguistics 2001
Period1/03/01 → …

Fingerprint

Context free grammars
Testing

Cite this

Belz, A. (2001). Optimisation of corpus-derived probabilistic grammars. In Corpus Linguistics 2001 (pp. 46-57). Lancaster, UK. https://doi.org/10.1.1.7.4428
Belz, Anja. / Optimisation of corpus-derived probabilistic grammars. Corpus Linguistics 2001. Lancaster, UK, 2001. pp. 46-57
@inproceedings{874eeeabbf59449788db37b4fe7839f1,
title = "Optimisation of corpus-derived probabilistic grammars",
abstract = "This paper examines the usefulness of corpus-derived probabilistic grammars as a basis for the automatic construction of grammars optimised for a given parsing task. Initially, a probabilistic context-free grammar (PCFG) is derived by a straightforward derivation technique from the Wall Street Journal (WSJ) Corpus, and a baseline is established by testing the resulting grammar on four different parsing tasks. In the first optimisation step, different kinds of local structural context (LSC) are incorporated into the basic PCFG. Improved parsing results demonstrate the usefulness of the added structural context information. In the second optimisation step, LSC-PCFGs are optimised in terms of grammar size and performance for a given parsing task. Tests show that significant improvements can be achieved by the method proposed. The structure of this paper is as follows. Section 2 discusses the practica",
author = "Anja Belz",
year = "2001",
month = "3",
day = "1",
doi = "10.1.1.7.4428",
language = "English",
pages = "46--57",
booktitle = "Corpus Linguistics 2001",

}

Belz, A 2001, Optimisation of corpus-derived probabilistic grammars. in Corpus Linguistics 2001. Lancaster, UK, pp. 46-57, Corpus Linguistics 2001, 1/03/01. https://doi.org/10.1.1.7.4428

Optimisation of corpus-derived probabilistic grammars. / Belz, Anja.

Corpus Linguistics 2001. Lancaster, UK, 2001. p. 46-57.

Research output: Chapter in Book/Conference proceeding with ISSN or ISBNConference contribution with ISSN or ISBNResearchpeer-review

TY - GEN

T1 - Optimisation of corpus-derived probabilistic grammars

AU - Belz, Anja

PY - 2001/3/1

Y1 - 2001/3/1

N2 - This paper examines the usefulness of corpus-derived probabilistic grammars as a basis for the automatic construction of grammars optimised for a given parsing task. Initially, a probabilistic context-free grammar (PCFG) is derived by a straightforward derivation technique from the Wall Street Journal (WSJ) Corpus, and a baseline is established by testing the resulting grammar on four different parsing tasks. In the first optimisation step, different kinds of local structural context (LSC) are incorporated into the basic PCFG. Improved parsing results demonstrate the usefulness of the added structural context information. In the second optimisation step, LSC-PCFGs are optimised in terms of grammar size and performance for a given parsing task. Tests show that significant improvements can be achieved by the method proposed. The structure of this paper is as follows. Section 2 discusses the practica

AB - This paper examines the usefulness of corpus-derived probabilistic grammars as a basis for the automatic construction of grammars optimised for a given parsing task. Initially, a probabilistic context-free grammar (PCFG) is derived by a straightforward derivation technique from the Wall Street Journal (WSJ) Corpus, and a baseline is established by testing the resulting grammar on four different parsing tasks. In the first optimisation step, different kinds of local structural context (LSC) are incorporated into the basic PCFG. Improved parsing results demonstrate the usefulness of the added structural context information. In the second optimisation step, LSC-PCFGs are optimised in terms of grammar size and performance for a given parsing task. Tests show that significant improvements can be achieved by the method proposed. The structure of this paper is as follows. Section 2 discusses the practica

U2 - 10.1.1.7.4428

DO - 10.1.1.7.4428

M3 - Conference contribution with ISSN or ISBN

SP - 46

EP - 57

BT - Corpus Linguistics 2001

CY - Lancaster, UK

ER -

Belz A. Optimisation of corpus-derived probabilistic grammars. In Corpus Linguistics 2001. Lancaster, UK. 2001. p. 46-57 https://doi.org/10.1.1.7.4428