Re: Rules WG -- draft charter -- NAF

   [Dan Conolly]
   log:includes corresponds quite nicely to the 'simple entailment'
   notion from the RDF semantics. i.e. F log:includes G iff
   G can be obtained from F by erasure and existential introduction.
   (F and G range over formulas).

   [me]
   > How is "entails" defined?

   In this case, as per RDF semantics. Existential-conjunctive logic.

But later ...

   >   We can't define entailment without a big fixed-point
   > construction of the sort beloved of nonmonotonic logicians.  Does the
   > CWM inference engine attempt to embody that construction?

   I'm not sure. I don't think so.

   > Of course we get apparent weirdness such as 
   > 
   > Fred's SW homepage:
   >             {Sally :homePage log:notIncludes {Fred a :Vegetarian}}
   > 	    => (Fred a :Carnetarian}
   > 
   > Sally's SW homepage:
   > 	    {Fred :homePage log:includes {Fred a :Carnetarian}}
   > 	    => {Fred a :Vegetarian}
   > 
   > where :Vegetarian and :Carnetarian are disjoint.  But we can just say
   > the two pages have multiple fixed points, and therefore imply nothing
   > about what Fred eats. 

   Yeah, I don't see much of a problem from that example.

It seems to me that you switch from one interpretation of "entails" to
another.  The first uses only RDF semantics (or an inference algorithm
that respects it).  The second one assumes that my conjectured
fixed-point algorithm would be correct.  If you apply the simpler one
to this example, then it seems inescapable that Fred's home page
asserts that Fred is a Carnetarian, because "Fred a :Vegetarian" is
not entailed (in the simple RDF way) from Sally's web page.  Sally's
web page doesn't assert anything, although it may at first glance
appear to be trying.

                                             -- Drew


-- 
                                   -- Drew McDermott
                                      Yale Computer Science Department

Received on Monday, 24 November 2003 21:55:20 UTC