Using WordNet to Extend FrameNet Coverage

Richard Johansson; Pierre Nugues

Using WordNet to Extend FrameNet Coverage

Research output: Chapter in Book/Report/Conference proceeding › Paper in conference proceeding › peer-review

169 Downloads (Pure)

Abstract

We present two methods to address the problem of sparsity in the FrameNet lexical database. The first method is based on the idea that a word that belongs to a frame is ``similar'' to the other words in that frame. We measure the similarity using a WordNet-based variant of the Lesk metric. The second method uses the sequence of synsets in WordNet hypernym trees as feature vectors that can be used to train a classifier to determine whether a word belongs to a frame or not. The extended dictionary produced by the second method was used in a system for FrameNet-based
semantic analysis and gave an improvement in recall.
We believe that the methods are useful for bootstrapping FrameNets for new languages.

Original language	English
Title of host publication	LU-CS-TR: 2007-240
Editors	Pierre Nugues, Richard Johansson
Publisher	Department of Computer Science, Lund University
Pages	27-30
Number of pages	4
ISBN (Print)	978-91-976939-0-5
Publication status	Published - 2007
Event	Building Frame Semantics Resources for Scandinavian and Baltic Languages - Tartu, Estonia Duration: 2007 May 24 → …

Conference

Conference	Building Frame Semantics Resources for Scandinavian and Baltic Languages
Country/Territory	Estonia
City	Tartu
Period	2007/05/24 → …

Subject classification (UKÄ)

Computer Sciences

Free keywords

Frame semantics
natural language processing
WordNet
FrameNet

Access to Document

johansson_nugues_frame2007

Cite this

@inproceedings{8fba37ed1c654f5d83112cec266dd392,

title = "Using WordNet to Extend FrameNet Coverage",

abstract = "We present two methods to address the problem of sparsity in the FrameNet lexical database. The first method is based on the idea that a word that belongs to a frame is ``similar'' to the other words in that frame. We measure the similarity using a WordNet-based variant of the Lesk metric. The second method uses the sequence of synsets in WordNet hypernym trees as feature vectors that can be used to train a classifier to determine whether a word belongs to a frame or not. The extended dictionary produced by the second method was used in a system for FrameNet-based semantic analysis and gave an improvement in recall. We believe that the methods are useful for bootstrapping FrameNets for new languages.",

keywords = "Frame semantics, natural language processing, WordNet, FrameNet",

author = "Richard Johansson and Pierre Nugues",

year = "2007",

language = "English",

isbn = "978-91-976939-0-5",

pages = "27--30",

editor = "Pierre Nugues and Richard Johansson",

booktitle = "LU-CS-TR: 2007-240",

publisher = "Department of Computer Science, Lund University",

note = "Building Frame Semantics Resources for Scandinavian and Baltic Languages ; Conference date: 24-05-2007",

}

TY - GEN

T1 - Using WordNet to Extend FrameNet Coverage

AU - Johansson, Richard

AU - Nugues, Pierre

PY - 2007

Y1 - 2007

N2 - We present two methods to address the problem of sparsity in the FrameNet lexical database. The first method is based on the idea that a word that belongs to a frame is ``similar'' to the other words in that frame. We measure the similarity using a WordNet-based variant of the Lesk metric. The second method uses the sequence of synsets in WordNet hypernym trees as feature vectors that can be used to train a classifier to determine whether a word belongs to a frame or not. The extended dictionary produced by the second method was used in a system for FrameNet-based semantic analysis and gave an improvement in recall. We believe that the methods are useful for bootstrapping FrameNets for new languages.

AB - We present two methods to address the problem of sparsity in the FrameNet lexical database. The first method is based on the idea that a word that belongs to a frame is ``similar'' to the other words in that frame. We measure the similarity using a WordNet-based variant of the Lesk metric. The second method uses the sequence of synsets in WordNet hypernym trees as feature vectors that can be used to train a classifier to determine whether a word belongs to a frame or not. The extended dictionary produced by the second method was used in a system for FrameNet-based semantic analysis and gave an improvement in recall. We believe that the methods are useful for bootstrapping FrameNets for new languages.

KW - Frame semantics

KW - natural language processing

KW - WordNet

KW - FrameNet

M3 - Paper in conference proceeding

SN - 978-91-976939-0-5

SP - 27

EP - 30

BT - LU-CS-TR: 2007-240

A2 - Nugues, Pierre

A2 - Johansson, Richard

PB - Department of Computer Science, Lund University

T2 - Building Frame Semantics Resources for Scandinavian and Baltic Languages

Y2 - 24 May 2007

ER -

Using WordNet to Extend FrameNet Coverage

Abstract

Conference

Subject classification (UKÄ)

Free keywords

Access to Document

Fingerprint

Cite this