Re: Expressing operators and functions in RDF

   [Stephane Fellah]
   Thank for your example and ontology. It is very interesting.  I would
   like to have opinions of the participants of this mailing list about how
   to express operators in RDF. I can distinguish two approaches: the CWM
   approach (relation as Property)
   (http://www.w3.org/2000/10/swap/doc/Built-In.html) versus your approach
   (relation as subject). Are they opposite/incompatible approach or is
   there any way to conciliate both ? 

The key difference between CWM and other approaches (Sintek & Decker's
TRIPLE system, our DRS, Horrocks and Patel-Schneider's OWL Rules) is
that CWM allows sets of triples to be labeled as internal to a rule.
Those triples are then not asserted.  You can't do that in RDF,
although perhaps it will become possible in RDF++, if there ever is
such a thing.

Hence the option of making a relation a property is simply not
available to any system but CWM.  If you tried saying "if triple1 then
triple2," the mere occurrence of triple1 and triple2 would cause them
to be asserted, so that "if p then q" would have the same meaning as
"p and q."

   Which one requires less parsing work ? 

I don't understand; parsing from what to what?  What kind of work --
programmer work or computer work?

   Mathematically, I think the CWM approach seems more correct. 

Mathematically the CWM approach _is_ more correct.  It's just not RDF, 
so it lives in a strange zone.  It seems more a vehicle for W3C people
to learn about forward and backward chaining than anything else.
Perhaps some day RDF will expand to make CWMism legal, at which point
hopefully other possibilities will become legal as well.

   Where can I find more resources discussing this issue ? 

If you explore the back alleys of W3C mail archives you can come
across discussions of the evolution of CWM features.  You can download
the latest CWM implementation (in Python) too.

   Is there any ontology developed using the CWM approach ?

I don't know.

-- 
                                             -- Drew McDermott
                                                Yale University CS Dept.

Received on Tuesday, 4 November 2003 13:29:30 UTC