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

Re: Should rdf:value have a semantics?? (was: Re: Quick review of RDFprimer)

From: Frank Manola <fmanola@mitre.org>
Date: Tue, 03 Dec 2002 15:00:45 -0500
Message-ID: <3DED0D6D.ABC5BD7C@mitre.org>
To: pat hayes <phayes@ai.uwf.edu>
CC: w3c-rdfcore-wg@w3.org

More discussion below:

pat hayes wrote:
> 
> >pat hayes wrote:
> >>
> >>  >
> >>  >General comment, not specifically Primer:  the description of
> >>  >rdf:value is fine, but how does it relate to a normative
> >>  >specification?  What can we say formally about rdf:value?
> >>
> >>  Right now, we say explicitly that it has no particular meaning.
> >>
> >>  >What formal semantics (interpretation) allows us to make inferences like:
> >>  >
> >>  >    my:cat rdf:type ex:DomesticCat .
> >>  >    my:cat ex:weight _:x .
> >>  >    _:x rdf:value "15" .
> >>  >    _:x ex:unit ex:Kilogram
> >>  >
> >>  >=>
> >>  >
> >>  >    my:cat rdf:type ex:Obese .
> >>
> >>  Nothing in the MT provides any connection in meaning between the
> >>  third and fourth triple of the first graph.
> >
> >I assume you mean beside the fact that they both have the same subject?
> 
> Well, yes. I mean, there isn't any formal support for that (or any
> other) entailment involving rdf:value
> 
> >Anyway, this isn't specific to rdf:value is it?
> 
> Well, it seems to me that rdf:value is being used in a way that kind
> of expects a semantics that it currently does not have. SO maybe we
> should give it one.
> 
> >I mean, you couldn't
> >make the inference if you'd used a datatype "weightInKilograms" either.
> >You'd need much more machinery than RDF has.
> 
> The point wasnt to do with the particular datatype, but the use of
> rdf:value. Right now rdf:value might as well not be there as far as
> the MT is concerned, it has no meaning at all. You could get the same
> effect by writing ex:foo instead.
> 
> >
> >>
> >>  I was under the impression that we had formally decided NOT to
> >>  support this kind of usage. Wasn't that part of the local-datatyping
> >>  no-fancy-idioms decision?
> >>
> >>  >but NOT:
> >>  >
> >>  >    my:cat rdf:type ex:DomesticCat .
> >>  >    my:cat ex:weight _:x .
> >>  >    _:x rdf:value "15" .
> >>  >    _:x ex:unit ex:Pound
> >>  >
> >>  >=>
> >>  >
> >>  >    my:cat rdf:type ex:Obese .
> >>  >
> >>  >?
> >>  >
> >>  >My point here is if we are to encourage such usage of rdf:value,
> >>  >then there ought to be some normative description to back up such
> >>  >usage.
> >>
> >>  I agree.  Either we should not mention this stuff, or else we should
> >>  back it up with some semantics.
> >
> >Well, that doesn't seem to be the principle on which we've been
> >operating!  If we're going to deal with this, let's do rdfs:isDefinedBy
> >too!
> 
> I would if I thought it had any meaning. The trouble with that one is
> that 'is defined by' really, really is meaningless when applied to
> pieces of RDF. RDF *never* defines anything, it can't possibly define
> anything since it doesn't have negation. So forget about
> rdfs:isDefinedBy, OK?

OK.

> 
> >
> >>  It wouldn't be hard to do and it
> >>  would also enable us to do some neat datatyping  entailments that
> >>  people seem to think are obvious, such as (with appropriate
> >>  wellformedness caveats)
> >>
> >>  aaa ppp "sss"^^ddd .
> >>
> >>  -->
> >>
> >>  aaa ppp _:x .
> >>  _:x rdf:value "sss" .
> >>  _:x rdf:type ddd .
> >>
> >
> >Yes, keeping in mind that rdf:value can be used with more complicated
> >relationships too (i.e., _:x could have multiple properties qualifying
> >the value, not just a single one).
> 
> Right, that's the limitation, but we can get that from the MT as well.
> 
> So more generally, to the WG: should I give rdf:value a model theory?
> Speak soon, guys. Unless I hear otherwise I will do this:
> 
> aaa rdf:value "bbb" .
> 
> means that there is some conventional mapping M from lexical forms
> under which I(aaa) = M(bbb). Doesn't say much, but it might be
> useful, particularly if we say that any datatype L2V mapping counts
> as a 'conventional mapping'.

I'm not clear on something.  This doesn't preclude the value of an
rdf:value property from being a typed literal does it?

--Frank
 


-- 
Frank Manola                   The MITRE Corporation
202 Burlington Road, MS A345   Bedford, MA 01730-1420
mailto:fmanola@mitre.org       voice: 781-271-8147   FAX: 781-271-8752
Received on Tuesday, 3 December 2002 15:01:04 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:54:47 EDT