RE: [SWC] RIF & OWL compatibility

Enrico and Jos,

yes, the issue of existential information is interesting
(and also in the field of databases there is no consensus
about null values being a good idea or not), but is it 
really an urgent issue for RIF? 

Some form of null values (or "blank nodes" or skolem 
constants/terms) may be useful, but probably not needed 
in phase 1. Do you know of any non-experimental (if not 
commercial) rule system that supports existential 
information?

The differences between a constructive LP-style 
interpretation of the existential quantifier and 
the non-consructive classical logic and DL-style 
interpretation seem to be even less relevant for
RIF than the issue itself.

-Gerd

--------------------------------------------
Gerd Wagner 
Brandenburg University of Technology 
at Cottbus, Germany







> -----Original Message-----
> From: public-rif-wg-request@w3.org 
> [mailto:public-rif-wg-request@w3.org] On Behalf Of Enrico Franconi
> Sent: Thursday, January 12, 2006 11:57 AM
> To: Jos de Bruijn
> Cc: public-rif-wg@w3.org
> Subject: Re: [SWC] Re: [OWL Compatibility] Re: RIF & OWL compatibility
> 
> 
> 
> On 12 Jan 2006, at 11:46, Jos de Bruijn wrote:
> >>> Wrt. the existential information: the semantics of LP rules is
> >>> based on
> >>> Herbrand models, which means (among other things) that every
> >>> individual
> >>> in the universe is represented with a name in the language. This  
> >>> makes
> >>> it impossible to truly capture existential information.
> >>
> >> I disagree. For example, In Rosati's approach non-distinguished
> >> varibales (i.e., existentials) are not restricted to the herbrand
> >> universe. BTW all the use cases in Managing Incomplete Information
> >> <http://www.w3.org/2005/rules/wg/wiki/
> >> Managing_incomplete_information> would be correctly handled by
> >> Rosati's approach.
> >
> > Rosati uses the standard names assumption, which excludes 
> considering
> > unnamed individuals, thus, certain kinds of existential knowledge  
> > cannot
> > be captured in his approach.
> 
> As in none LP based approaches.
> 
> > If you have a name for the individual, then you can certainly deal  
> > with
> > it in the logic program.
> > However, if we would take the following example:
> >
> > take the following first-order sentence:
> > \exists x. p(x)
> >
> > and the following rule:
> >
> > r <- p(x)
> >
> > One cannot conclude r, because there is no name a such that p(a).
> 
> In Eiter, the above rule would not derive r, and in Rosati it  
> couldn't even be written since it is not DL-safe. You could  
> automatically make it DL-safe, and then you wouldn't derive r either.
> 
> However, unlike Eiter, and unlike pure LP based approaches, in his  
> recent work Rosati does allow for several kind of classical  
> existentials. For example, the use cases involving OWL-Lite in  
> <http://www.w3.org/2005/rules/wg/wiki/ 
> Managing_incomplete_information> would be captured correctly by  
> Rosati but not by Eiter.
> 
> cheers
> --e.
>

Received on Thursday, 12 January 2006 13:36:40 UTC