W3C home > Mailing lists > Public > public-esw-thes@w3.org > October 2006

Re: SKOS Core enquiry

From: Mark van Assem <mark@cs.vu.nl>
Date: Tue, 24 Oct 2006 12:11:57 +0200
Message-ID: <453DE6ED.6040005@cs.vu.nl>
To: Dan Brickley <danbri@danbri.org>
CC: Quentin Reul <qreul@csd.abdn.ac.uk>, public-esw-thes@w3.org


Hi,

>>See [1] for the currently completed draft. There are no plans for SKOS
>>versions, although I did include some ideas about that. You have to make
>>a choice if you want to see Synsets or WordSenses as skos:Concepts.
> 
> 
> Or both?

Possibly. Either choice has pros and cons. If you want both:

wn:Synset 			-> skos:Concept
	[all its lexical
	forms]			-> skos:altLabel
	
wn:WordSense 			-> skos:Concept
	[wordform attached	
	 to the Sense's Word]	-> skos:prefLabel

wn:hyponymOf			-> skos:broader
wn:[other props between synsets]-> skos:related

left over: props between WordSenses, and a missing link between 
Synsets and WordSenses. All Synset-concepts are "orphans" (not related 
to the WordSense-concepts with a skos relation, while the 
Wordsense-concepts carry the hierarchical info). This results in a bit 
strange thesaurus but doesn't loose that much info.

But it's maybe a better idea to drop the WordSenses altogether.
Synsets form the real idea of a "concept" in WordNet.
But then you have to drop the relations between wordsenses, and you 
have to group all hyponym relations of a synset's wordsenses into the 
new skos concept (and then turn them into skos:broader)? Then you 
really start reinterpreting the original source...

Mark.

-- 
  Mark F.J. van Assem - Vrije Universiteit Amsterdam
        markREMOVE@cs.vu.nl - http://www.cs.vu.nl/~mark
Received on Tuesday, 24 October 2006 10:12:27 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:38:55 GMT