Re: rdfs:isDefinedBy and version IRI

I think I remember hearing at SemTech the other week that rdfs:isDefinedBy
has fallen out of favor. I'm not sure if that is because people just have
not adopted it or because the working group nudged us not to use it.

I guess I'll try to answer your question with another question. What are the
variety of uses for rdfs:isDefinedBy according to
1) as intended, and 2) in practice?

The answers that I know of:
* Provide a mechanism for an automated agent to retrieve the ontology for a
term that it doesn't know (enough) about.
* Provide a triple-based string comparison in a query over terms (instead of
performing a string regex in a FILTER to limit only predicates within a
certain namespace, simply adding another triple in a query ( $subject $p $o
.. $p rdfs:isDefinedBy myns: ) .

That's all I have...

Regards,
Tim Lebo


On Sat, Jun 13, 2009 at 4:41 AM, Simon Reinhardt
<simon.reinhardt@koeln.de>wrote:

> Hello,
>
> When writing OWL ontologies in RDF syntaxes it is generally a good practice
> to explicitly relate the defined terms (i.e. classes, properties and
> individuals) back to the ontology through the property "rdfs:isDefinedBy".
> In OWL 2 with the introduction of owl:versionIRI there are two options now:
>
> - Relate the terms to the ontology IRI
> - Relate the terms to the version IRI
>
> I wonder if the working group wishes to state any preference or give advice
> on which option to choose?
>
> Regards,
>  Simon
>
>