Re: Could owl:sameAs reference non-OWL resources?

[This is a reply to the posting of this message on www-rdf-logic.]

From: "Bernard Vatant" <bernard.vatant@mondeca.com>
Subject: Could owl:sameAs reference non-OWL resources?
Date: Mon, 28 Apr 2003 12:14:06 +0200

> Somehow a follow-up of my previous question about "things" ...
> 
> I'm currently trying to figure how OWL can interoperate with legacy in
> other languages and formats, in particular DAML libraries, Topic Maps
> Published Subject Indicators, and in general any format using URIs to
> name-define "things" (or RDF resources, or Topic Maps subjects). Could be
> as well plain on-line HTML thesaurus or glossary, Dewey Classes, LoC
> Subject Headings ...
> 
> Seems to me that one important aim of OWL (the main one?) should be to use
> those URIs to achieve the "colocation objective", which is also the main
> objective of Topic Maps and Published Subjects.
> 
> The colocation objective in the distributed Web environment means being
> able to assert that two "things" are identical. owl:sameAs allows to
> achieve that it seems. But what is not clear to me is if the URI referenced
> by owl:sameAs has to define an OWL resource, or if it could be another kind
> of resource.

I'm not sure what you are trying to get at with the phrase ``OWL
resource''.   

OWL makes the assumption that all individuals (i.e., entities that are
suitable for use with properties like owl:sameAs) belong to the class
owl:Thing.  However, there are very few restrictions on what can be an
instance of owl:Thing.  In OWL DL (and OWL Lite) the only restriction is
that instances of owl:Thing can't be classes or properties or structural
elements (like rdf:nil).  In OWL Full there are no restrictions
whatsoever. 

So if you are in OWL Full there are essentially no limitations on the
applicability of owl:sameAs.

> In the example given in http://www.w3.org/TR/owl-ref/
> 
> <owl:Class rdf:ID="FootballTeam">
>   <owl:sameAs rdf:resource="http://sports.org/US#SoccerTeam"/>
> </owl:Class>
> 
> Is "http://sports.org/US#SoccerTeam" supposed to reference an OWL element?

Well, yes, in that anything, in OWL Full, or just about anything, in OWL DL
and OWL Lite, can be ``an OWL element''.

> What if it reference a DAML legacy for instance?

No problem, except, of course, that what you probably want is to include
the DAML+OIL meaning for any DAML+OIL ontologies that concern this entity,
which you are not going to get.

> Or an XTM topic element? 

Ditto.

> Should it be converted into OWL before being used that way? 

Well, as stated above, any DAML+OIL or XTM ontologies/knowledge
bases/whatever will not be understood by an OWL-only system, and thus
probably should be converted.

> IOW, will owl:sameAs ensure backward compatibility with DAML+OIL
> libraries? Or support XTM subject indicators? Or UDDI, XRI ... whatever.

No.  At least not unless someone builds a system that can handle both OWL
and DAML+OIL or XTM or UDDI, or whatever.  

Why would you expect anything different.

> Otherwise said, will one be able to use owl:sameAs with the same
> flexibility as XTM <subjectIndicatorRef> which has basically the same
> semantics? The above example would read in XTM :
> 
> <topic id="FootballTeam">
> 	<subjectIdentity>
> 		<subjectIndicatorRef xlink:href="http://sports.org/US#SoccerTeam"/>
> 	<subjectIdentity>
> </topic>
> 
> <subjectIndicatorRef> does not assume any particular format for the
> resource at "http://sports.org/US#SoccerTeam". Could be an XML fragment,
> could be HTML, could be OWL or any other relevant syntax. The
> recommendation on Published Subjects somehow give clues on what kind of
> resource should be used that way. BTW this recommendation is currently in
> the same final draft status than OWL. See
> http://www.ontopia.net/tmp/pubsubj-gentle-intro.htm
> 
> Bottom line: OWL elements can be used as Published Subject Indicators,
> provided the OWL ontology contains some specific information (like
> publisher identity, human-readable definitions ...) that can be expressed
> in OWL. So OWL should be IMO (the most) recommended format for Published
> Subject Indicators. Please note that so far it's only a personal view, that
> I will try to push of course on the table of Published Subjects TC ASAP
> (next week).
> But to achieve the round trip, it would be cool if through owl:sameAs, one
> could reference a Published Subject Indicator using e.g. XTM syntax. Being
> able to ensure that round trip would be IMO a giant step towards semantic
> integration.

What round trip?  You appear to be asking for a system that can handle
multiple formalisms.  This is not a trivial thing, and can often be very
problematic, or even impossible.

> Feedback much welcome.
> 
> 
> Bernard Vatant
> Senior Consultant
> Knowledge Engineering
> Mondeca - www.mondeca.com
> bernard.vatant@mondeca.com


Peter F. Patel-Schneider
Bell Labs Research
Lucent Technologies

Received on Monday, 28 April 2003 07:59:43 UTC