RE: exchanging OWL through RIF

---- Original message ----
>Date: Fri, 3 Mar 2006 09:49:40 -0800
>From: "Vincent, Paul D" <PaulVincent@fairisaac.com>  
>Subject: RE: exchanging OWL through RIF   
>To: <bparsia@isr.umd.edu>, "RIF WG" <public-rif-wg@w3.org>
>
>At the very least this (Hoolet) is a "reasoning engine", determining (new 
logical) associations from (prior ontological, FOL?) statements.

Hoolet will determine entailments of SWRL kbs. SWRL is a superset of OWL DL 
and a subset of FOl.

>The Rete-type of rule engine is more of a "decisioning engine" and expects to 
work against data.

I have no idea what you mean. My post was to point out some systems that 
Michael Kifer either didn't know about or don't meet his criteria for non-on-
paper FOL rule engine.

Regardless, Hoolet can work "against data". A subset of RDF is a subset of SWRL 
(and OWL DL). (Call it the non-hilog subset of RDF/S).

> Sure, that data could be association information, but typically it is not the "is 
my brothers wife's brother my relative" type query (/rule).

Er...even if so? So what?

The examples on the Hoolet page are not supposed to be exemplars...just 
examples.

>I am still trying to work out how these worlds join (in a practical, useful way)...  

Since I, frankly, have no idea what you are talking about, I'm afraid I can't help. 

My point was simply that there are implementors who are targeting SWRL (and 
various subsets like AL Log or DL Safe rules). Of course, these systems don't 
have the history of e.g., CLIPS or JESS, but neither are they merely paper.

(KAON2 has an implemention of DL Safe rules, which are FOL. And is 
commercially available. Hoolet is closer to a toy or experimental system. Pellet is 
a pretty robust and reasonable system, but the rule support is very preliminary 
(i.e., first cut, naive implementation of DL Safe rules should be completed next 
week). Disclaimer, I work on Pellet. Pellet does have some AL Log support.)

Cheers,
Bijan.

Received on Friday, 3 March 2006 18:44:09 UTC