[ACT-R-users] WN-Lexical question
Bruce J Weimer MD
bjweimer at charter.net
Wed Jun 18 11:41:05 EDT 2008
Re: [ACT-R-users] WN-Lexical questionBruno,
Thank you for the explanation and code! I suspected that there was a certain amount of randomness inherent in this system. But I have a question - there are several definitions for "dog"... but if I ask you to define "dog", you would almost certainly pick:
(a member of the genus Canis (probably descended from the common wolf) that has been domesticated by man since prehistoric times; occurs in many breeds)
which is the first definition that appears when you search WordNet on-line:
http://wordnet.princeton.edu/perl/webwn?s=dog&o2=&o0=1&o7=&o5=&o1=1&o6=&o4=&o3=&h=
In fact, the WordNet on-line responses seem to ordered according to the most common meanings first. I was just wondering if we could get at the definitions ranked according to usage......... it seems that somehow they do..........
Bruce.
---- Original Message -----
From: Emond, Bruno
To: Bruce J Weimer MD ; ACT-R
Sent: Tuesday, June 17, 2008 9:41 AM
Subject: Re: [ACT-R-users] WN-Lexical question
Bruce,
In principle, there is no first or last gloss in the lexicon.
There is just some higher probability that some lexical elements will be retrieved given a prior context.
In Wn-Lexical, a random selection, a set-difference or a set intersection constraint can be specified in the retrieval request.
Also, in Wordnet every lexical entry and operator has a synset-id attached to it. Words having the same synset-ids are synonyms.
This way you can have a constrained retrieval of a specific gloss or word given a known synset-id.
I have attached to this email a new set of files for the WNLexical module.
In particular, you could have a look at the model find-all-synomyns.lisp as an example of getting a specific word sense.
More elaborate models could use the hyponyms and hypernyms operators.
Bruno
On 6/15/08 7:56 PM, "Bruce J Weimer MD" <bjweimer at charter.net> wrote:
I'm looking at the example "wnl-find-definition.lisp" - it returns a random gloss for the word... is there a way to get it to return the first gloss?
Bruce.
----------------------------------------------------------------------------
_______________________________________________
ACT-R-users mailing list
ACT-R-users at act-r.psy.cmu.edu
http://act-r.psy.cmu.edu/mailman/listinfo/act-r-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/act-r-users/attachments/20080618/90c3a773/attachment.html>
More information about the ACT-R-users
mailing list