W3C home > Mailing lists > Public > www-rdf-logic@w3.org > April 2004

RE: types of OWL

From: Dan Connolly <connolly@w3.org>
Date: Fri, 02 Apr 2004 14:33:26 -0600
To: ewallace@cme.nist.gov
Cc: www-rdf-logic@w3.org
Message-Id: <1080938006.16761.2067.camel@dirk>

On Fri, 2004-04-02 at 14:09, ewallace@cme.nist.gov wrote:
> Jeremy Carroll wrote:
> 
> 
> >> "OWL Lite" goes
> >> further in
> >> that direction, by ommiting some constructs known to be tough to work with
> >> using DL techniques.
> >>
> >
> >From a reasoner's point of view OWL Lite is not that much lighter than OWL
> >DL (in fact some of the hardest of the OWL Test Cases are in the 'harder OWL
> >Lite' section, where I was perverse to give implementors a challenge).
> >
> >The distinguishing feature that makes Lite, Lite (in my view) is that from
> >the point of a view of a *person* trying to understand (or write) an
> >ontology it is easier (unless people have been perverse, and expressed
> >ontologies which conceptually should be in OWL DL, but can be coded up into
> >OWL Lite).
> 
> I respectfully disagree.
> 
> The design principle we agreed to for OWL Lite was ease of implementation.

We agreed on exactly what our records say we agreed on.

I find both ease of implementation and the point of view of
persons reading/writing ontologies in the recorded motivation for
OWL Lite:

"OWL Lite supports those users primarily needing a classification
hierarchy and simple constraints. For example, while it supports
cardinality constraints, it only permits cardinality values of 0 or 1.
It should be simpler to provide tool support for OWL Lite than its more
expressive relatives, and OWL Lite provides a quick migration path for
thesauri and other taxonomies. Owl Lite also has a lower formal
complexity than OWL DL"
  -- http://www.w3.org/TR/2004/REC-owl-features-20040210/#s1.3



FYI, the relevant issue was
http://www.w3.org/2001/sw/WebOnt/webont-issues.html#I5.2-Language-Compliance-Levels

but there is no rationale included in the group's decision to close it:

RESOLVED: to close 5.2, endorsing existing an owl lite language subset
and test class
  -- http://www.w3.org/2001/sw/WebOnt/ftf4


(perhaps you meant design goal, rather than design principle,
but ease of implementation is not among our design goals either
http://www.w3.org/TR/webont-req/#section-goals )


> In the end though, the difference between DL and Lite really only seems 
> significant for DL reasoners.  Lite is not a significantly simpler language 
> to learn, and I can't see how maxCardinality 2 is any harder to understand or 
> write then maxCardinality 0 or 1.  These days, I don't even think about Lite, 
> but I am often asking myself if my OWL is still DL.
> 
> -Evan
-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/
see you at the WWW2004 in NY 17-22 May?
Received on Friday, 2 April 2004 15:33:39 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:52:48 GMT