- From: Stephen Reed <reed@cyc.com>
- Date: Wed, 29 May 2002 11:46:35 -0500 (CDT)
- To: "R.V.Guha" <guha@guha.com>
- cc: Bill Andersen <andersen@ontologyworks.com>, RDF Logic <www-rdf-logic@w3.org>
Given the intentional limitations of DAML+OIL with regard to logical
expresiveness, the relevant portion of the Cyc deductive inference engine
is what we call Subsumption Based Heuristic Layer (SBHL). The internal
representation is a graph of the subsumption relationships in the DAML
class, instance and property objects, enabling deduction to proceed in the
same manner as I understand description logic implementations.
Cyc's explanation of the inference chain for a typical DAML inference
might be:
Q: Does a dog have fur?
A: Yes, because dogs are a subclass of mammal, and mammals have the fur
property.
Q: Does a plant have fur?
A: No, because the class of plants is disjoint with the class of animals,
and the class of animals are the only things I know of having the fur
property.
Cyc need not be singled out for particular discussion given the numerous
subsumption-based RDF-triple deductive inference implementations, because
the higher order, more complex inference methods (for example
backchaining) do not get used in a strict DAML+OIL representation.
Regards to Cyc contributors on this list.
-Steve
On Thu, 23 May 2002, R.V.Guha wrote:
> I did not say that Cyc has a model theory. Nor did I say that it is predictable in its current form. However, I
> do believe that it can be made predictable, etc. with good software engineering, i.e., without a model theory.
--
===========================================================
Stephen L. Reed phone: 512.342.4036
Cycorp, Suite 100 fax: 512.342.4040
3721 Executive Center Drive email: reed@cyc.com
Austin, TX 78731 web: http://www.cyc.com
download OpenCyc at http://www.opencyc.org
===========================================================
Received on Wednesday, 29 May 2002 12:47:41 UTC