W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > October 2002

Re: incomplete entailments [Was: email straw poll: literal semantics proposals]

From: Patrick Stickler <patrick.stickler@nokia.com>
Date: Thu, 10 Oct 2002 18:29:27 +0300
Message-ID: <000e01c27071$d213e610$544516ac@NOE.Nokia.com>
To: "Frank Manola <fmanola" <fmanola@mitre.org>, "ext Jos De_Roo" <jos.deroo.jd@belgium.agfa.com>
Cc: "RDF Core" <w3c-rdfcore-wg@w3.org>, <w3c-rdfcore-wg-request@w3.org>, "brian_mcbride" <brian_mcbride@hp.com>



[Patrick Stickler, Nokia/Finland, (+358 40) 801 9690, patrick.stickler@nokia.com]


----- Original Message ----- 
From: "ext Jos De_Roo" <jos.deroo.jd@belgium.agfa.com>
To: "Frank Manola <fmanola" <fmanola@mitre.org>
Cc: "RDF Core" <w3c-rdfcore-wg@w3.org>; <w3c-rdfcore-wg-request@w3.org>; "brian_mcbride" <brian_mcbride@hp.com>
Sent: 10 October, 2002 17:35
Subject: incomplete entailments [Was: email straw poll: literal semantics proposals]


> 
> 
> Frank, it seems to me a primitive entailment
> and as such a basic and core entailment for
> a machine to make, but indeed not like NLP
> (natural language processing)
> Avoiding to make it is like having something
> like a hidden C, D or E

Bingo! That's precisely it. F provides an implicit C
like interpretation of the inlined literals.

Thus, for any literal -- typed or inlined -- the interpretation 
is based on 

   DATATYPE + LEXICAL_FORM = VALUE

and if you have both in the node label, such as with a 
typed literal node, then you have all the info you need
to get to the value (with knowledge of the datatype of 
course). I.e.

   xsd:integer + "10" = 10

If you only have the lexical form in the node label, such
as an inlined literal, then you need the "missing" datatype 
part from somewhere else to determine its meaning, such as
provided by an rdfs:range constraint.

   ??? + "10" = ???

   ... rdfs:range xsd:integer .

   xsd:integer + "10" = 10 

But in either case, the node is taken to denote a datatype 
value and comparisons of two inlined literal nodes simply
cannot be done if the datatype information is missing.

Think of it like a function that takes four arguments, 
the datatype and lexical form of the first node and the
datatype and lexical form of the second node, and returns
true or false. E.g.

   equalValues(dt1, lf1, dt2, lf2)

if you're missing the datatype for one or both of the literals,
then you can't pass all the required arguments to the function
for comparision. E.g.

   equalValues(null, "LLL", null, "LLL")

and the comparison simply cannot be made.

Eh?

Patrick


>                     Frank Manola                                                                                   
>                     <fmanola@mitre       To:     Jos De_Roo <jos.deroo.jd@belgium.agfa.com>                        
>                     .org>                cc:     "Brian McBride <bwm", RDF Core <w3c-rdfcore-wg@w3.org>,           
>                                           w3c-rdfcore-wg-request@w3.org                                            
>                     2002-10-10           Subject:     Re: email straw poll: literal semantics proposals            
>                     02:39 PM                                                                                       
>                                                                                                                    
>                                                                                                                    
> 
> 
> 
> 
> Jos--
> 
> Can you say a little more about why you don't want to lose this
> entailment?  Note: my understanding is that even under F an application
> would still be able to access the triples
> 
>         :I :love "RDF".
>         :You :love "RDF".
> 
> and deduce anything it wanted, including
> 
>         :I :love _:x.
>         :You :love _:x.
> 
> (presumably on the basis that the two "RDF" strings *may* have been
> meant to refer to the same thing).  It's only entailments directly
> defined by the RDF semantics we're talking about here, not additional
> determinations that applications may want to make, right?
> 
> --Frank
> 
> Jos De_Roo wrote:
> 
> >
> > [sorry to be late, but I had no access in Bristol to my mailbox]
> >
> > B 5 (for rdf:format xsd:... there are canonical lexical forms)
> > C 3 (should we then always say e.g. owl:cardinality xsd:integer"2"?)
> > D 2 (meaning is same as C, but less kiss)
> > F 0 i.e. I do not want to loose the the entailment
> >       :I :love _:x.
> >       :You :love _:x.
> >     given *nothing* but
> >       :I :love "RDF".
> >       :You :love "RDF".
> >     (ref. meaningful derivation requirement)
> >
> 
> 
> --
> Frank Manola                   The MITRE Corporation
> 202 Burlington Road, MS A345   Bedford, MA 01730-1420
> mailto:fmanola@mitre.org       voice: 781-271-8147   FAX: 781-271-875
> 
> 
> 
> 
> 
> 
Received on Thursday, 10 October 2002 11:29:30 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:52:23 EDT