W3C home > Mailing lists > Public > public-owl-wg@w3.org > April 2008

Re: Profiles intro

From: Bijan Parsia <bparsia@cs.man.ac.uk>
Date: Thu, 10 Apr 2008 13:25:40 +0100
Message-Id: <FD53BD3D-158B-4DD5-AFBA-CB1EE63FB0E7@cs.man.ac.uk>
Cc: Carsten Lutz <clu@tcs.inf.tu-dresden.de>, OWL Working Group WG <public-owl-wg@w3.org>
To: Ivan Herman <ivan@w3.org>

On 10 Apr 2008, at 12:36, Ivan Herman wrote:

> Hi Carsten,
[snip]
> I am trying to find what the _emphasis_ is, and not to be  
> exhaustive! That is why I used the word 'emphasis'. What would you  
> put in place instead?

This is part of the trick, yes? It's hard to be emphatic without  
suggesting exhaustion.

[snip]
> Ouch. That hurts. This is a Semantic Web ontology, so if an  
> ontology is unrelated to RDF triplets, than what does it have to do  
> with this group?

Surely Semantic Web != RDF or even, necessarily, RDF like. For  
example, consider the homogeneity of RDF syntax (everything is a  
triple). That's not shared by other web languages (e.g., HTML or SVG,  
nor is it shared by many  other KR languages. Arguably (though you  
may disagree) it's an accidental feature of RDF rather than an  
essential feature. OWL-R is designed, afaik, to be triple oriented.

Furthermore, I would be surprised if the W3C thought that members had  
to commit *at all* to the web aspects of OWL in order to be a  
participant in the group or to advocate features or profiles. Surely,  
the W3C wants to be responsive to member needs?

> I do not see why OWL-R would be more RDF-ish than DL Lite or vice  
> versa.

RDF, in practice, is existential free (this is what we argue about  
with bnodes). RDF rule implementations tend to skolemize and not  
infer Bnodes as variables. OWL-R is designed not to have existentials  
in the head. Thus, it is more RDF-ish.

> We are talking about data and, possibly, lots of it. Those are  
> typically  RDF or RDF-able on the Semantic Web.

But it's not the *Semantic Web* ontology language, it's the *Web*  
ontology language!

(Sorry, couldn't resist ;))

> That DL-Lite may be of interest outside of the Semantic Web may be  
> true, but is besides the point in this environment...

Exactly not, I would say.  Intranets and walled gardens are important  
to HTML. Non semantic web and even non web uses of OWL are in scope.  
(Dan Connolly often makes this point.) After all, the more contexts  
OWL is useful, the more likely it'll get good support.

>>> profiles are defined: DL-Lite, that can be implemented on top of  
>>> traditional database systems using query rewriting, and OWL-R,  
>>> that can be implemented using basic rule systems.
>>> ]]]
>> True.
>>> I hope this is at least factually correct. It is interesting to  
>>> note that on such high level there is no real difference between  
>>> DL-Lite and OWL-R, and the only way to differentiate them on that  
>>> level is how they are implemented.
>> Yes, a main difference between DL-Lite and OWL-R is implementation
>> techniques. Another one is maybe RDF-ishness.
>
> where I strongly disagree.

Do you disagree that OWL Full is more RDF(S)ish than OWL DL? It seems  
a relatively harmless observation.

I do think that DL Lite is generally fairly RDF(S)ish. E.g., when I  
first saw it I immediately though that it was a nigh-perfect small  
superset of RDFS. But surely it isn't out of court to point out that  
it's a bit less triplely, per se?

[snip]
> On the other hand, we need names that are suggestive and helpful to  
> people. Having said that, 'OWL DL' is a meaningless name for those  
> who do not know what description logic is...

Not really. I mean, there are plenty of people who just know it as  
OWL DL as in "y'know, that think pellet implements". Oversuggestion  
can be dangerous (as I said before).

Cheers,
Bijan.
Received on Thursday, 10 April 2008 12:24:05 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 10 April 2008 12:24:06 GMT