RE: Are DeprecatedClasses invisible to DIG Reasoners?

Hi, Bill...
    I also have little experience in interpreting these documents, but
I'll give it a shot.  I don't know what additional issues the DIG
protocol might bring into play, so I can only comment on the OWL
aspects.  It appears that you might be getting a little confused between
the semantics of a _class_ that is _typed_ as a DeprecatedClass and the
actual DeprecatedClass _axiom_, itself.  I believe it is the axiom that
"has no meaning in the model theoretic semantics other than that given
by the RDF(S) model theory", meaning that an OWL reasoner should not
treat the class any differently than if it was not typed as a
DeprecatedClass.
    This interpretation of the statement you quoted from the OWL
Reference appears to be in line with the semantics you cited from
section 3.3 of the OWL Semantics document, wherein the presence of a
DeprecatedClass axiom only expands the extension of the rdf:type
relation and does not otherwise affect the semantics.
    Given this, I believe you could say that DeprecatedClasses are NOT
invisible to reasoners, but that the DeprecatedClass axiom (effectively)
is.  (Mind you, I'd say this is true for OWL DL, though not so much for
OWL Full, about whose semantics I will not claim any solid
understanding.)  I believe this is in line with most models for
deprecation in computer languages.  That is, something which is
deprecated will continue to operate as it always, but the user of it (be
it a programmer, modeler, user, etc.) may receive warning messages to
remind them that they should try to transition away from using it.

-Brandon :)


 


________________________________

From: public-owl-dev-request@w3.org
[mailto:public-owl-dev-request@w3.org] On Behalf Of William Bug
Sent: Saturday, August 18, 2007 5:33 PM
To: public-owl-dev@w3.org
Subject: Are DeprecatedClasses invisible to DIG Reasoners?


Hi All, 

This is a naive question from someone who's never - in the context of
implementing a DIG Reasoner - had to interpret either the OWL Direct
Model-Theoretic Semantics:
http://www.w3.org/TR/2004/REC-owl-semantics-20040210/direct.html#owl_Dep
recatedClass_semantics
or the OWL Abstract Syntax:
http://www.w3.org/TR/2004/REC-owl-semantics-20040210/syntax.html#owl_Dep
recatedClass_syntax_lite
for owl:DeprecatedClass.

I've read the W3C OWL docs - I've searched this lists archive - I've
read the SWOOP paper that talks a bit about versioning - and I've
Googled about, but nowhere can I find a simple answer to the question:

Are DeprecatedClasses invisible to DIG Reasoners?

The following sentence in the OWL Language Reference implies they might
be:

"(DeprecatedClass) has no meaning in the model theoretic semantics other
than that given by the RDF(S) model theory."
http://www.w3.org/TR/2004/REC-owl-ref-20040210/#Deprecation

I could write a simple OWL file with logical entailments from which a
reasoner can identify inferred sets, then go in and declare some of the
Classes as DeprecatedClasses to see how this effects the reasoner
output, but I thought I'd simply come to the source and ask the experts.
Something tells me from the variety of scenarios a reasoner might have
to confront when parsing a complex OWL file with specific Classes and/or
Properties defined as deprecated that the answer may not be
straight-forward.

Having said this, given in OWL DL, owl:Class is a direct subclass of
rdfs:Class (as opposed to being equivalent to rdfs:Class) and
owl:DeprecatedClass is also a direct subclass of rdfs:Class, it seems
reasonable to assume DeprecatedClasses definitions have not logical
entailments.

What would it mean, then, for a Class definition to contain the
following type specification:
<rdf:type rdf:resource="http://www.w3.org/2002/07/owl#DeprecatedClass"/>
I've noticed when I specify an owl:Class as "Deprecated" in Protege-OWL
(<= v3.3beta), this element is added to the otherwise unchanged class
specification.

Many thanks for any info or citations you can provide on this issue.

Cheers
Bill Bug


Bill Bug
Senior Research Analyst/Ontological Engineer

Laboratory for Bioimaging  & Anatomical Informatics
www.neuroterrain.org
Department of Neurobiology & Anatomy
Drexel University College of Medicine
2900 Queen Lane
Philadelphia, PA    19129
215 991 8430 (ph)
610 457 0443 (mobile)
215 843 9367 (fax)


Please Note: I now have a new email - William.Bug@DrexelMed.edu