- From: Sandro Hawke <sandro@w3.org>
- Date: Thu, 10 Apr 2008 09:44:06 -0400
- To: "Peter F. Patel-Schneider" <pfps@research.bell-labs.com>
- Cc: ivan@w3.org, bparsia@cs.man.ac.uk, clu@tcs.inf.tu-dresden.de, public-owl-wg@w3.org
Peter writes:
> It seems to me very much germane to discriminate between the profiles
> based on their position related to the stance that everything is a
> triple. Of the three profiles, OWL-R is the only one that has a version
> that is "Full", i.e., uniformly treats all information as triples. Both
> EL++ and DL-Lite distinguish between data (in the form of triples) and
> other kinds of information (namely ontology information).
>
> It thus seems more accurate to say that of the three profiles OWL-R is
> the only one that can indiscriminately handle triples without losing its
> fundamental characteristics. Both EL++ and DL-Lite are completely
> agnostic as to whether the underlying data model of non-fact axioms is
> based on triples and could not deal with indiscriminate mixing of fact
> triples and ontology triples.
>
> If I was trying to say this in as few words as possible I would probably
> have said that OWL-R is the profile most closely tied to RDF, or even
> that OWL-R is the only profile that is very RDF-ish.
I think that last bit goes too far. I think a significant fraction of
RDF users (I would guess 40% now, but climbing to 95% in ten years) stay
outside of the pointy-brackets. They don't mess with the classes and
predicates; they just use them. Those folks are using RDF, but not in a
way that distinguishes between DL and Full.
More-over, I think it's important that our specs say people SHOULD stay
away situations where DL and Full differences matter.
-- Sandro
Received on Thursday, 10 April 2008 13:45:21 UTC