Re: Concept Equivalence, IFPs, skos:subjectIndicator and owl:sameAs (was Re: SKOS Guide and owl:sameAs)

Bernard Vatant wrote:
> Hi Alistair
> 
> Happy to see you have a look at this. Mieux vaut tard que jamais :-)
> 
> Since the thread has been quite long, do you want me to try and sum up 
> the issue?

I would be very grateful if you could.

In no more than three sentences? ;)

Thanks,

Alistair.


> 
> Cheers
> 
> Bernard
> 
> Alistair Miles a écrit :
>>
>> Hi Stuart,
>>
>> Quick comment without having read the subsequent thread in detail ...
>>
>> I think you have revealed a potential inconsistency in the design of 
>> SKOS. Certainly worthy of an item in the issues list - I'll do that 
>> when I get a chance.
>>
>> Cheers,
>>
>> Alistair.
>>
>> Williams, Stuart (HP Labs, Bristol) wrote:
>>> Hello,
>>>
>>> Firstly I am new to SKOS, so apologies if this has come up before.
>>>
>>> The SKOS guide (as indicated in this thread) discourages the use of 
>>> owl:sameAs to establish equivalence relations between skos:Concepts 
>>> because:
>>>
>>> <http://www.w3.org/TR/swbp-skos-core-guide/#secidentity>
>>> "The property owl:sameAs should not be used to express the fact that 
>>> two conceptual resources (i.e. resources of type skos:Concept) share 
>>> the same meaning. The property owl:sameAs implies that two resources 
>>> are identical in every way (they are in fact the same resource). 
>>> Although two conceptual resources may have the same meaning, they may 
>>> have different owners, different labels, different documentation, 
>>> different history, and of course a different future."
>>>
>>> However, the use of skos:subjectIndicator, defined inverse 
>>> functional, is described in:
>>>
>>> <http://www.w3.org/TR/swbp-skos-core-guide/#secpsi>
>>> "The property skos:subjectIndicator allows you to assert a link 
>>> between a concept and a human-readable document that provides a 
>>> complete, definitive description of that concept.
>>> ...
>>> The skos:subjectIndicator property is an 
>>> owl:InverseFunctionalProperty, which means that if any two nodes in 
>>> an RDF graph have the same value for this property, then they are the 
>>> same resource [OWL]. Therefore you can use the skos:subjectIndicator 
>>> property to indirectly identify a concept by reference to the URI of 
>>> the document that is the published subject indicator for that concept."
>>>
>>> So... if the same skos:subjectIndicator is asserted for two (or more) 
>>> skos:Concepts then it can be inferred that they are the same 
>>> skos:Concept which seems to be at odds with the desire *not* to 
>>> establish such equivalences as expressed in the narrative above 
>>> discouraging the use of owl:sameAs.
>>>
>>> Is there a reason why concept equivalences established via the 
>>> skos:subjectIndicator are "good" and equivalencies established via 
>>> owl:sameAs are bad?
>>>
>>> Thanks,
>>>
>>> Stuart Williams
>>> -- 
>>> HP Labs, Bristol.
>>>
>>>> Hi,
>>>>
>>>> There is a confusion of different things here.
>>>>
>>>> 1- skos:Concept is a class.
>>>> 2- instances of skos:Concept are called "concepts" in the SKOS 
>>>> documents.
>>>>
>>>> 3- mapping of skos:Concept to another class can be done with 
>>>> owl:equivalentClass
>>>> 4- mapping of instances of skos:Concept to other skos:Concept 
>>>> instances (from other vocabularies) can be done with owl:sameAs
>>>> 5- mapping of instances of skos:Concept to other instances (from 
>>>> other vocabularies) can also be done with the SKOS mapping 
>>>> properties, e.g. exactMatch [1]
>>>>
>>>> Now the confusion is about which kind of mapping (3-5) is meant. The 
>>>> "Concept Identity and Mapping" section [2] states that mapping type 
>>>> 4 should not be used, instead type 5 is better. This is because the 
>>>> former states that they are the same *in every respect*, while the 
>>>> latter only states that their extensions are the same (set of docs 
>>>> indexed with one concept is also properly indexed with the other). 
>>>> If you use the former you also merge their metadata, e.g. date of 
>>>> creation and scheme they belong to. They become indistinguishable. 
>>>> The latter keeps them distinguishable.
>>>>
>>>> The text mentioned does not refer to type 3 at all. This mapping 
>>>> would be required if someone is not using the SKOS schema for a 
>>>> vocabulary, but  something similar. Then a mapping
>>>>
>>>>     skos:Concept owl:equivalentClass my:Concept
>>>>
>>>> can be used to make all instances of my:Concept also skos:Concepts, 
>>>> so they can be manipulated by software that understands SKOS.
>>>>
>>>> Hope this helps,
>>>> Mark.
>>>>
>>>> [1]http://www.w3.org/2004/02/skos/mapping/spec/#exactMatch
>>>> [2]http://www.w3.org/TR/2005/WD-swbp-skos-core-guide-20051102/#secidentity 
>>>>
>>>>
>>>> Nabonita Guha wrote:
>>>>> */Bernard Vatant <bernard.vatant@mondeca.com>/* wrote:
>>>>>
>>>>>  >A skos:Concept is not a class, and the domain of 
>>>>> owl:equivalentClass is
>>>>>  >owl:Class
>>>>>
>>>>> Whereas in SKOS Core guide 
>>>>> (http://www.w3.org/TR/2005/WD-swbp-skos-core-guide-20051102/#secconcept), 
>>>>> skos:Concept has been described as a class. If it's not a Class 
>>>>> then what it can be considered as?
>>>>>
>>>>> Best,
>>>>>
>>>>> Nabonita Guha
>>>>>
>>>>> Senior Research Fellow
>>>>> Documentation Research & Training Centre
>>>>> Indian Statistical Institute
>>>>> Bangalore INDIA
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>     Hi Andrew
>>>>>      > The SKOS guide [1], in the "Concept Identity and Mapping"
>>>>>     section, states that owl:sameAs *should not* be used to indicate
>>>>>     that two concepts share the same meaning. It gives some 
>>>>> rationale in
>>>>>     the section for this. Looking at the OWL guide [2], in the "4.1.
>>>>>     Equivalence between Classes and Properties" section, I'm wondering
>>>>>     whether one can use owl:equivalentClass to indicate that two
>>>>>     concepts share the same meaning. If there is a reason why
>>>>>     owl:equivalentClass can/cannot be used for this, should it also be
>>>>>     mentioned in "Concept Identity and Mapping" section?
>>>>>      >
>>>>>     Yes, there is a good reason.
>>>>>     A skos:Concept is not a class, and the domain of 
>>>>> owl:equivalentClass is
>>>>>     owl:Class
>>>>>
>>>>>     Cheers
>>>>>
>>>>>     Bernard
>>>>>
>>>>>      >
>>>>>      > Thanks, Andy.
>>>>>      >
>>>>>      > [1]
>>>>>      > [2]
>>>>>      >
>>>>>      >
>>>>>      >
>>>>>
>>>>>     --
>>>>>     *Bernard Vatant
>>>>>     *Knowledge Engineering
>>>>>     ----------------------------------------------------
>>>>>     *Mondeca**
>>>>>     *3, cité Nollez 75018 Paris France
>>>>>     Web: www.mondeca.com
>>>>>     ----------------------------------------------------
>>>>>     Tel: +33 (0) 871 488 459
>>>>>     Mail: bernard.vatant@mondeca.com
>>>>>     Blog: Leçons de Choses
>>>>>
>>>>>
>>>>>
>>>>> ------------------------------------------------------------------------ 
>>>>>
>>>>> Want to start your own business? Learn how on Yahoo! Small 
>>>>> Business. 
>>>>> <http://us.rd.yahoo.com/evt=41244/*http://smallbusiness.yahoo.com/r-index> 
>>>>>
>>>> -- 
>>>>   Mark F.J. van Assem - Vrije Universiteit Amsterdam
>>>>         markREMOVE@cs.vu.nl - http://www.cs.vu.nl/~mark
>>>
>>
> 

-- 
Alistair Miles
Research Associate
CCLRC - Rutherford Appleton Laboratory
Building R1 Room 1.60
Fermi Avenue
Chilton
Didcot
Oxfordshire OX11 0QX
United Kingdom
Web: http://purl.org/net/aliman
Email: a.j.miles@rl.ac.uk
Tel: +44 (0)1235 445440

Received on Monday, 27 November 2006 17:10:27 UTC