Re: A triple is not unique.

Sergey wrote:
> That's my interpretation:

> - every statement has a unique ID generated using a Skolem function
>   reify(s, p, o).
> - the reification mechanism described in the specification (or any other
>   we come up with) is syntactic;
>   the parser will substitute the value of reify(s, p, o) for the
>   identifier used in the serialization.

> A triple is unique.

Wonderful, Sergey! (how are you doing?)
Let my try it also ...
We should proof that:
  Forall s,p,o Exists t: t is unique [F]
Skolem has found that we can eliminate exists and write:
  Forall s,p,o: f(s,p,o) is unique [Fs]
Formula Fs is not equivalent with formula F, but:
  F is provable <=> Fs is provable
Now, a function has a unique value for its arguments, so
  QED

This also shows the good news about implicit universal quantification
and the use of proof-value RESOURCE's for RDF based logic ...

--
Jos De Roo, AGFA http://www.agfa.com/w3c/jdroo/

Received on Tuesday, 21 November 2000 17:44:27 UTC