Re: A Model Theoretic Semantics for DAML-ONT (now, an Axiomatic Semantics)

This message is to announce a new version of our document that specifies
a formal semantics for the DAML-ONT language.
The new document is available in 3 forms:

word:      www.ksl.stanford.edu/people/dlm/DAML-Ont-kif-axioms-001113.doc
html:        www.ksl.stanford.edu/people/dlm/DAML-Ont-kif-axioms-001113.html
and text:   www.ksl.stanford.edu/people/dlm/DAML-Ont-kif-axioms-001113.txt

also, our DAML project page is being kept updated with new release pointers:
www.ksl.stanford.edu/projects/daml/

The new document contains revisions that respond to the comments we
received on the initial version of the document.  In particular, this
version makes clear that what is being presented is an axiomatic
semantics for DAML-ONT and a prescription for producing a logical theory
in KIF that is logically equivalent to a DAML-ONT ontology.  Since KIF
has a model theoretic semantics and the document specifies how to
translate a DAML-ONT ontology into an equivalent KIF logical theory, we
have in effect specified a model theoretic semantics for DAML-ONT.
However, as was pointed out in several comments, the document itself
does not directly specify a model theoretic semantics for DAML-ONT.

There has been some discussion as to why we chose KIF as the language in
which to axiomatize DAML-ONT.  That is basically a design choice on our
part.  Our primary rationale for using KIF was to make the axiomatization as
simple and easy to understand as possible for the broadest audience.
The resulting axioms require assuming little more than conventional first
order logic.  They, for example, do not require the assumption of a set
theory in addition to first order logic.  The footnotes are intended to
provide a reader who is familiar with predicate logic with enough
information about KIF to understand the axioms.  There will no doubt be
multiple versions of DAML-ONT axiomatizations in multiple representation
languages as the DAML community grows and matures.  We hope that this
one helps that growth and maturation by making the language easier to
understand and to use.   After we have reached consensus on this one, we are willing to
provide a set theoretic semantics as well.

Thanks to all those who provided feedback, particularly Pat Hayes.  Your reviews and insights have been quite useful.  More recent comments from Jeff Heflin and Dan Connolly will be responded to in our next update.

As stated in the document, comments are welcome posted to the
www-rdf-logic@w3.org distribution list.

Richard and Deborah

=======================
Deborah L. McGuinness
Associate Director and Senior Research Scientist
Knowledge Systems Laboratory
Stanford University
Stanford, CA 94305
dlm@ksl.stanford.edu
voice  650 723 9770
fax 650 725 5850

Received on Tuesday, 14 November 2000 02:33:32 UTC