W3C home > Mailing lists > Public > public-esw-thes@w3.org > April 2007

Re: SKOS properties

From: Guus Schreiber <schreiber@cs.vu.nl>
Date: Mon, 30 Apr 2007 00:48:33 +0200
Message-ID: <463520C1.2090206@cs.vu.nl>
To: Bernard Vatant <bernard.vatant@mondeca.com>
CC: Sue Ellen Wright <sellenwright@gmail.com>, Stella Dextre Clarke <sdclarke@lukehouse.demon.co.uk>, Quentin Reul <qreul@csd.abdn.ac.uk>, SWD Working Group <public-swd-wg@w3.org>, public-esw-thes@w3.org



Bernard Vatant wrote:
> 
> Hi Sue Ellen and all
> 
> I will keep agnostic, to begin with, on the question of knowing if 
> antonymy as the dark side of synonymy. But I would like to point that in 
> any case, technically it does not make sense to use "owl:disjointWith" 
> property to link two skos:Concept(s), simply because a skos:Concept is/ 
> en principe/ not a class (in any case not an owl:Class), and 
> owl:disjointWith is used to link two owl:Class to express that they have 
> no common instance. Of course in OWL-Full nothing can prevent you to 
> declare that a skos:Concept is also a owl:Class, but the logical 
> consequences of such a declaration are unpredictable :-)

In OWL Full skos:Concept IS an owl:Class 
(rdfs:Class and owl:Class are equivalent in OWL 
Full). I also fail to see the damage you could do 
with a owl:disjointWith statement.

Guus


> 
> If one wants to use owl:disjointWith for what I guess Quentin and you 
> have in mind, and make it in a clean way, one should define in OWL the 
> class of all resources indexed by some "skos:Concept", using a 
> "owl:hasValue" restriction on "skos:subject", and then declare that the 
> class of resources (documents) with subject "white" is disjoint with the 
> class of resources with subject "black". And I'm pretty sure this is not 
> true, so I tend to balance rather on Stella's side. But I'm reluctant to 
> go as far as declaring those two classes as "owl:equivalentClass", which 
> would be the logical expression of considering "white" and "black" as 
> synonyms. But certainly the intersection is not empty : many, if not all 
> resources with subject "black" have also the subject "white" (IMO). So 
> if the classes are not equivalent, they are definitely not disjoint.
> 
> So ... I don't know. As Stella says, the standards "allow you" to admit 
> antonyms as some kind of synonyms/equivalents, or rather to consider a 
> pair of antonyms as two faces of the same concept. But do they 
> "recommend" it? And BTW in the case of "black" and "white", on which 
> basis should I choose "black" rather than "white" as preferred, and the 
> other as synonym? ( ... too hard an issue for 1.15 a.m.)
> 
> Bernard
> 
> Sue Ellen Wright a écrit :
>> Hi, All,
>> As a terminologist, the notion of adding antonyms as 
>> equivalents/synonyms strikes me as really undesirable. In an 
>> ontology-like environment it would really be problematic. By the same 
>> token, it is hard to classify antonym relations -- this has long been 
>> a subject of debate in terminology/lexicography circles. I rather like 
>> the idea of "disjointwith" together with a scope note. Especially in 
>> multilingual concept management, knowing the antonym is often a real 
>> clue to the disambiguation of the concept associated with a term.
>> Bye for now
>> Sue Ellen
>>
>>  
>> On 4/26/07, *Stella Dextre Clarke* <sdclarke@lukehouse.demon.co.uk 
>> <mailto:sdclarke@lukehouse.demon.co.uk>> wrote:
>>
>>     You may like to know that ISO 2788 and BS 8723 both allow you to
>>     admit antonyms as though they were equivalents (with relationship
>>     tagged USE/UF) if appropriate. For example, in my own thesaurus I
>>     have an entry "Inconsistency of indexing USE Indexing consistency"
>>     because both of these terms are actually referring to the same
>>     underlying concept. (A scope note might describe it  as "the
>>     degree of  consistency or inconsistency encountered in indexing".)
>>     If you want to be more precise, you could set it up as a special
>>     type of equivalence relationship.
>>          SKOS could choose to handle antonyms the same way, if it wishes.
>>     (*some* antonyms, I should stress - not all examples would be
>>     suitable for this treatment.) In an ontology, you might prefer the
>>     relationships to be more specific.
>>          Cheers
>>     Stella
>>          *****************************************************
>>     Stella Dextre Clarke
>>     Information Consultant
>>     Luke House, West Hendred, Wantage, Oxon, OX12 8RR, UK
>>     Tel: 01235-833-298
>>     Fax: 01235-863-298
>>     SDClarke@LukeHouse.demon.co.uk 
>> <mailto:SDClarke@LukeHouse.demon.co.uk>
>>     *****************************************************
>>
>>         -----Original Message-----
>>         *From:* public-esw-thes-request@w3.org
>>         <mailto:public-esw-thes-request@w3.org>
>>         [mailto:public-esw-thes-request@w3.org
>>         <mailto:public-esw-thes-request@w3.org>] *On Behalf Of
>>         *Quentin Reul
>>         *Sent:* 26 April 2007 12:08
>>         *To:* SWD Working Group
>>         *Cc:* public-esw-thes@w3.org <mailto:public-esw-thes@w3.org>
>>         *Subject:* SKOS properties
>>
>>         Hi all,
>>         I was looking at the properties available as part of SKOS and
>>         realized that there wasn't any properties to represent
>>         antonyms. However, these are sometimes useful and present in
>>         some thesauri such as WordNet. Would owl:disjointWith be
>>         sufficient to represent antonyms?
>>         Thanks,
>>         Quentin
>>
>>         --         
>> ------------------------------------------------------------------------
>>
>>         Quentin H. Reul
>>         Computing Science
>>         University of Aberdeen
>>
>>         +44 (0)1224 27 *4485*
>>         qreul@csd.abdn.ac.uk <mailto:qreul@csd.abdn.ac.uk>
>>         http://www.csd.abdn.ac.uk/~qreul
>>         <http://www.csd.abdn.ac.uk/%7Eqreul>
>>
>>
>>
>>
>> -- 
>> Sue Ellen Wright
>> Institute for Applied Linguistics
>> Kent State University
>> Kent OH 44242 USA
>> sellenwright@gmail.com <mailto:sellenwright@gmail.com>
>> swright@kent.edu <mailto:swright@kent.edu>
>> sewright@neo.rr.com <mailto:sewright@neo.rr.com>
>> ------------------------------------------------------------------------
>>
>> No virus found in this incoming message.
>> Checked by AVG Free Edition. Version: 7.5.463 / Virus Database: 
>> 269.6.0/775 - Release Date: 24/04/2007 17:43
>>   
> 

-- 
Vrije Universiteit Amsterdam, Computer Science
De Boelelaan 1081a, 1081 HV Amsterdam, The Netherlands
T: +31 20 598 7739/7718; F: +31 84 712 1446
Home page: http://www.cs.vu.nl/~guus/
Received on Sunday, 29 April 2007 22:48:54 GMT

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