RE: Core concepts and relationships as an ontology?

I agree with Mike. It would be a good exercise and one we would
encourage--however, we are using UML and Prose for formality.
As these modes of expression seem to be more understood by our
members, I think this where we need to focus while enrouraging
others to help with alternative expressions.

DaveH


-----Original Message-----
From: Champion, Mike [mailto:Mike.Champion@SoftwareAG-USA.com]
Sent: Tuesday, June 24, 2003 7:12 AM
To: www-ws-arch@w3.org
Subject: RE: Core concepts and relationships as an ontology?





> -----Original Message-----
> From: Peter Mika [mailto:pmika@cs.vu.nl]
> Sent: Tuesday, June 24, 2003 8:38 AM
> To: www-ws-arch@w3.org
> Subject: Core concepts and relationships as an ontology?
> 

> I was wondering if the Core Concepts and Relationships part of the
> latest WSA document is available anywhere as an ontology, 
> preferably in
> RDF(S) or DAML+OIL?

As Katia is on holiday or other duties IIRC, I'm not sure who else we have
who is well-versed in these technologies.  FWIW, I hade a Gestalt moment at
the Phoenix F2F when I realized that the Concepts/Relationships section is
an "ontology" and I would be happy if we could (perhaps as an appendix)
publish a formal description of the WS architecture in some appropriate
formal language.  

I think the core working group is (and should remain) pretty focused on
defining the core "ontology" in *English*, but it would be very interesting
to hear from people with the time and ability to propose specific languages
and tools that we could use to formalize the model as it evolves.  It's even
possible [hint, hint, if this Semantic Web stuff is as cool as it is
supposed to be!] that ontology concepts and tools would make the job easier
than it is using slippery English. Most of us are more familiar with UML as
a formalism to get around the imprecision of natural language, but I for one
would be glad to be enlightened about the state of the art in the ontology
world if that could replace or supplement our UML efforts with something
more appropriate to the level of abstraction at which we are working.

Thoughts?  I'm personally a bit worried that spending much effort in this
area could get us into the familiar morass of debating what the meaning of
"is" is, and so on.  On the other hand, I'm a bit desperate to see some of
this mass of Jello get nicely nailed onto trees, and if using formal
ontologies helps, I'm willing to learn!

Received on Tuesday, 24 June 2003 12:02:35 UTC