RE: definitionOf

> The document http://www.w3.org/TR/2002/WD-rdf-mt-20021112/
> says in Appendix A, "RDF Axioms":
>     rdf:type(?x,?y)  iff  ?y(?x)
> or in my alternate KR syntax
>     $x has rdf:type = $y  iff  $x isa* $y
> $x was not restricted to Individuals.
> Note: my "isa" is the "union" of "isu" and "iss", i.e., it is
> valid for $x which is either an individual or a species.

How do you seperate individuals from species, without regard to a particular
context?

> 2. expressing "definitionOf" in triples
> Deferred.
> Remarks:
> a. I identified "definitionOf" as a necessary property to describe
> reality -- the human method of concept-formation.
> b. "definitionOf" is a ternary property.
> c. Is the purpose of OWL to describe reality, or to see what can be
> done with binary properties?

a. Whether this cannot currently be done may require analysis (I am neither
qualified to do so, nor in any temper to attempt, but we obviously need more
than one person looking at the question).

b & c. The purpose of OWL is surely to describe reality *in a way that is of
practical use*. Triples have practical advantages; small units can be dealt
with by computers efficiently, particularly in distributed environments, if
at the cost of occasional verbosity for somethings that would more naturally
be a single statement. Besides which the triples don't come from OWL, the
are there already in the underlying RDF layer - OWL can't change that.

My question here is mainly practical. There are a couple of different
obvious ways of expressing a ternary relationship in triples, I was
inquiring as to which you'd suggest.

Received on Tuesday, 19 November 2002 12:19:50 UTC