RE: OWL-cool

At 09:52 PM 4/6/2003 +0100, Peter Crowther wrote:

>snip
> > When you
> > buy into OWL-DL, you may be kissing good-bye to
> > rule-based inference.
>
>Why is this bad?  Given the availability of DL inference engines such as
>FaCT and RACER, and experience with the integration of these with rules
>engines in systems such as TRIPLE, I'd argue that we can benefit from the
>advantages of both.  I don't see why we should shackle ourselves to logical
>systems that are amenable to rule-based inference.

No one that I've run into claims to have a DL system that also understands
rules, or a rule system that also understands DL.  If someone does, I'd
like to hear about it.  Without that, there is no synergy between the
DL portion of such a hybrid ontology and the rule part, in which case
you are missing out on any inferences that touch both sides (and if
there isn't any crossover, why would the two have been combined?).

Cheers, Bob

Received on Monday, 7 April 2003 23:10:44 UTC