I have just finished reviewing the
referenced OMG document, specifically chapters 9 and parts of 10. This is
not what I was referring to. This document seems intended to generate the
mappings between various approaches, WITHOUT developing a group consensus metamodel
of what is or can be in an ontology
I used italics because I
understand, as would anyone who has read the definition as adopted in chapter 9
of this document that there is a very wide variety of opinion as to what
constitutes an ontology. Therefore I say again, break out into the most
elemental, i.e. atomic parts those constituents of ontology, and create an
inclusive, abstract metamodel describing the many relationships/constraints
existing between these elemetns in whatever language is preferred. [Because of
the wide variety of perspectives involved, I am sure that many elements would
be hotly contested as to their inclusion/exclusion and these could be treated
with appropriate caveats in the metamodel.] Then we can debate to endless
delight as to what constitutes the completeness, richness etc.; and then
we can map each of our candidates to this [agreed on metamodel] in terms of
what the candidate contains, and what it does not.
Jacob A. Teller
Senior
Scientist
NCI Information Systems, Inc.
10010
Junction Drive, Suite 202,
Annapolis
Junction, MSD 20701
jteller@xxxxxxxxx 301 939-4404
-----Original Message-----
From:
ontology-summit-bounces@xxxxxxxxxxxxxxxx [mailto:ontology-summit-bounces@xxxxxxxxxxxxxxxx]
On Behalf Of Cory Casanave
Sent: Thursday, January 18, 2007
6:31 PM
To: Ontology Summit 2007 Forum
Subject: Re: [ontology-summit]
categorizing "ontology"
ODM (http://www.omg.org/docs/ad/06-05-01.pdf)
does a good job of providing meta models for multiple ontology languages (such
as CL, OWL, Topic Maps) as well as mappings between many of them.
It does not, however, normalize those meta models into common elements or
"atomic structures" and, as Bill says, does not say anything about
the capabilities or use cases for those languages.
-Cory Casanave
From: ontology-summit-bounces@xxxxxxxxxxxxxxxx
[mailto:ontology-summit-bounces@xxxxxxxxxxxxxxxx] On Behalf Of Bill Andersen
Sent: Thursday, January 18, 2007
3:26 PM
To: Ontology Summit 2007 Forum
Subject: Re: [ontology-summit]
categorizing "ontology"
This exists and it's called ODM, or at least until
something better comes along. But ODM really only says things about
documents formed from sentences in this or that logical language and doesn't
inform the discussion of "ontology" as meant by Steve, Peter, and
Leo.
On Jan 18, 2007, at 14:31 , Teller, Jacob wrote:
Suppose there existed a meta model
of “Ontology” that created and linked all of the atomic structures
that an ontology could/should have according to
the consensus of the Ontolog community. Then this meta-model would be
capable of having mappings to all of those local “Ontologies” that
are in reality thesauri, or conceptual models, or taxonomies, or use cases,
etc. These mappings would clearly show what is and IS NOT included in the
local offering and this then becomes a relatively straightforward set of
comparison points for what one is referring to as “I have an
Ontology.”
Jacob A. Teller
Senior Scientist
NCI Information Systems, Inc.
10010 Junction Drive, Suite 202,
Annapolis Junction, MSD 20701
jteller@xxxxxxxxx 301 939-4404
This email
transmission contains information from NCI Information Systems, Inc. that may
be considered privileged or confidential and is intended solely for the named
recipient. If you have received this message in error, please contact the
sender immediately and be aware that the use, copying or dissemination of this
information is prohibited.
_________________________________________________________________