Inducing Compact but Accurate Tree-Substitution Grammars

Trevor Cohn; Sharon Goldwater; Phil Blunsom

Inducing Compact but Accurate Tree-Substitution Grammars

Trevor Cohn, Sharon Goldwater, Phil Blunsom

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract / Description of output

Tree substitution grammars (TSGs) are a compelling alternative to context-free grammars for modelling syntax. However, many popular techniques for estimating weighted TSGs (under the moniker of Data Oriented Parsing) suffer from the problems of inconsistency and over-fitting. We present a theoretically principled model which solves these problems using a Bayesian non-parametric formulation. Our model learns compact and simple grammars, uncovering latent linguistic structures (e.g., verb subcategorisation), and in doing so far out-performs a standard PCFG.

Original language	English
Title of host publication	Proceedings of Human Language Technologies: The 2009 Annual Conference of the North American Chapter of the Association for Computational Linguistics
Place of Publication	Boulder, Colorado
Publisher	Association for Computational Linguistics
Pages	548-556
Number of pages	9
ISBN (Print)	978-1-932432-41-1
Publication status	Published - 1 Jun 2009

Access to Document

N09 1062Final published version, 172 KB

Cite this

@inproceedings{6989ad95c7e8481fa5948b3102cc707b,

title = "Inducing Compact but Accurate Tree-Substitution Grammars",

abstract = "Tree substitution grammars (TSGs) are a compelling alternative to context-free grammars for modelling syntax. However, many popular techniques for estimating weighted TSGs (under the moniker of Data Oriented Parsing) suffer from the problems of inconsistency and over-fitting. We present a theoretically principled model which solves these problems using a Bayesian non-parametric formulation. Our model learns compact and simple grammars, uncovering latent linguistic structures (e.g., verb subcategorisation), and in doing so far out-performs a standard PCFG.",

author = "Trevor Cohn and Sharon Goldwater and Phil Blunsom",

year = "2009",

month = jun,

day = "1",

language = "English",

isbn = "978-1-932432-41-1",

pages = "548--556",

booktitle = "Proceedings of Human Language Technologies: The 2009 Annual Conference of the North American Chapter of the Association for Computational Linguistics",

publisher = "Association for Computational Linguistics",

}

Cohn, T, Goldwater, S & Blunsom, P 2009, Inducing Compact but Accurate Tree-Substitution Grammars. in Proceedings of Human Language Technologies: The 2009 Annual Conference of the North American Chapter of the Association for Computational Linguistics. Association for Computational Linguistics, Boulder, Colorado, pp. 548-556. <http://www.aclweb.org/anthology/N/N09/N09-1062>

Inducing Compact but Accurate Tree-Substitution Grammars. / Cohn, Trevor; Goldwater, Sharon; Blunsom, Phil.
Proceedings of Human Language Technologies: The 2009 Annual Conference of the North American Chapter of the Association for Computational Linguistics. Boulder, Colorado: Association for Computational Linguistics, 2009. p. 548-556.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Inducing Compact but Accurate Tree-Substitution Grammars

AU - Cohn, Trevor

AU - Goldwater, Sharon

AU - Blunsom, Phil

PY - 2009/6/1

Y1 - 2009/6/1

N2 - Tree substitution grammars (TSGs) are a compelling alternative to context-free grammars for modelling syntax. However, many popular techniques for estimating weighted TSGs (under the moniker of Data Oriented Parsing) suffer from the problems of inconsistency and over-fitting. We present a theoretically principled model which solves these problems using a Bayesian non-parametric formulation. Our model learns compact and simple grammars, uncovering latent linguistic structures (e.g., verb subcategorisation), and in doing so far out-performs a standard PCFG.

AB - Tree substitution grammars (TSGs) are a compelling alternative to context-free grammars for modelling syntax. However, many popular techniques for estimating weighted TSGs (under the moniker of Data Oriented Parsing) suffer from the problems of inconsistency and over-fitting. We present a theoretically principled model which solves these problems using a Bayesian non-parametric formulation. Our model learns compact and simple grammars, uncovering latent linguistic structures (e.g., verb subcategorisation), and in doing so far out-performs a standard PCFG.

M3 - Conference contribution

SN - 978-1-932432-41-1

SP - 548

EP - 556

BT - Proceedings of Human Language Technologies: The 2009 Annual Conference of the North American Chapter of the Association for Computational Linguistics

PB - Association for Computational Linguistics

CY - Boulder, Colorado

ER -

Inducing Compact but Accurate Tree-Substitution Grammars

Abstract / Description of output

Access to Document

Fingerprint

Cite this