Errors caused by word sense ambiguity
To: firstname.lastname@example.org, email@example.com
Subject: Errors caused by word sense ambiguity
From: Adam Kilgarriff <firstname.lastname@example.org>
Date: Tue, 25 Apr 95 17:56:21 BST
Old-Received: from nora.hd.uib.no by alfie.uib.no with SMTP (PP) id <email@example.com>; Tue, 25 Apr 1995 18:57:36 +0200
Old-Received: from sun3.nsfnet-relay.ac.uk by nora.hd.uib.no with SMTP id AA07785 (5.65c8/IDA-1.4.4 for <firstname.lastname@example.org>); Tue, 25 Apr 1995 19:00:24 +0200
Old-Received: from mail-relay.ja.net by sun3.nsfnet-relay.ac.uk with local SMTP id <email@example.com>; Tue, 25 Apr 1995 17:57:03 +0100
Old-Via: uk.ac.brighton.it-research-institute; Tue, 25 Apr 1995 17:56:50 +0100
Resent-Date: Tue, 25 Apr 1995 18:57:57 +0200
Request for accounts of errors caused by word sense ambiguity
Do you have an NLP application which needs to establish the meaning
of the input?
Has the application ever made mistakes because it thought it had one
sense of a word, when in fact it had another? (And if not why not!?)
If so, I'd be very interested in hearing from you...
While there is now a substantial literature on the problem of word
sense disambiguation, this is almost always divorced from any
application (1). The goal is to disambiguate between the senses given
in a dictionary or thesaurus on the grounds that that, or something
similar, is necessary for full understanding (and is an interesting
problem in its own right). If this work is to feed in to NLP
applications, we need to ask: where, and how, does word sense
ambiguity cause problems for applications? We can then address
questions of how disambiguation work can be made more practical and
how it can be customised to particular applications. This is one goal
of SEAL, our EPSRC-funded grant at the University of Brighton.
As a first step, I'm gathering examples and anecdotes, as well as
references, papers, or figures if you have any, of the sorts of
problems that ambiguity has actually caused.
Please note that I am not concerned with word class ambiguity (eg
'bank' as a verb or a noun) but only ambiguity within a word class
(nominal 'bank' as money-bank or river-bank).
All comments and contributions most welcome!
(1) For a notable exception, see Dagan and Itai, Computational
Linguistics 20(4) (1994), who describe a word sense disambiguation
method specifically tailored to machine translation.
Adam Kilgarriff tel: (44) 1273 642919
Research Fellow (44) 1273 642900
Information Technology Research Institute fax: (44) 1273 606653
University of Brighton
Lewes Road email:
Brighton BN2 4AT firstname.lastname@example.org