Re: A basis for convergence and closure?

> >Pat,
> >
> >we have tried to summarize that
> >at http://www.agfa.com/w3c/euler/rdfd-theory.n3
> >is that making sense ???
>
> Is that 'a' in the first triple in the antecedent shorthand for
> rdf:type?? If so that seems to all make sense, except the first
> triple in the antecedents ought to be

yes, ``a'' is shorthand for ``rdf:type''

> ?d a rdf:DataType .
>
> Ie d is in the class of datatypes.  dType is the datatyping property.

of course
I've also added
  rdf:DataType a rdfs:Class .

>   Of course this doesn't actually state the literal-interpretation
> rules, but I don't think you can say those in RDF.
>
> I'd prefer to state the second rule consequent as
>
> ?d rdfs:subPropertyOf rdf:value .

well, I must say that I don't yet see that yet
but that's maybe because I better go and sleep now :-)

> but I can see that would give you a much longer inference path. Your
> version ought to work, though, in every case I can think of :-)

I'm curious to experiment with that...

> Pat

> PS On a different topic, did Ora contact you about the closure rules?
> He has some startling statistics on how redundant they are.

we had a couple of messages indeed
(but I now start to realize that point)

--
Jos

Received on Wednesday, 6 February 2002 19:34:06 UTC