Re: RDF speficiations (was RE: Cutting the Patrician datatype knot)

From: "Jonathan Borden" <jborden@mediaone.net>
Subject: Re: RDF speficiations (was RE: Cutting the Patrician datatype knot)
Date: Mon, 3 Dec 2001 18:32:58 -0500

> Peter F. Patel-Schneider wrote:

[...]

> Right and that's why it is important to be totally clear about how XML
> Schema might interface with RDF. My fear is that slapping this together
> might cause subtle unintentioned problems (so far, I've been extremely
> disappointed in the lack of coordination among various W3 WGs including RDF,
> XSD, XLink/XPointer, XHTML etc.)
> 
> >
> > Sure, and so, maybe, RDF is not in compliance with XML.  If so then this
> is
> > a problem that should be worked out in the Semantic Web Coordination
> Group.
> 
> he he, it is clear that coordination with various XML related activities is
> not often the highest priority.

Yes, and I find this disturbing and mystifying.  If the RDF Core WG is not
taking care of this then some other body within W3C should.  It appears to
me that this sort of coordination is THE purpose of the Semantic Web
Coordination Group, at least from my reading of

	http://www.w3.org/2001/sw/CG/Overview.html

and

	http://www.w3.org/2001/sw/CGcharter.html

However, I don't see any action from them, nor do I see any consideration
of this issue by them, based on my reading of the member-available
information about the Semantic Web Coordination Group.

[...]

> > If you define myxsd:integer as something that defines a lexical-to-value
> > mapping in a different way than xsd:integer does, then, of course any
> > implementation of RDF plus datatypes that understands whatever method you
> > used to define myxsd:integer must follow your definition.  If the value
> > space of myxsd:integer is the same as the value space of xsd:integer and
> > the lexical-to-value mapping is the usual one, then, again of course, the
> > RDF plus datatypes implementation is *required* to answer affirmatively
> > that <John> <age> <myxsd:integer:010> entails <John> <age>
> <xsd:integer:10>
> > Otherwise the implementation is non-conforming.  How can it be otherwise?
> 
> Aha! But do you assume that the requirements you have outlined above are
> implemented by a particular XML Schema validation interface?

Yes.  In fact, the XQuery Data Model supports precisely this interface, or
so I gather from my reading of the specification.

> This is my entire point, even though it is quite apparent that two types are
> equal (their instance sets are identical), discovery of such type
> equivalence is not necessarily possible by a type system that uses
> _hierarchy alone_ in determining type algebra operations i.e. are two types
> equal, supersets or subsets. An _explicit_ type heirarchy tells you this
> "because I say so", but more generally a type formalism allows such type
> equivalences to be computed from the equations that define the types
> themselves.

Sure, and so you need access to an implementation of XML Schema datatypes,
such as will be (or is) provided by an implementation of the XQuery Data
Model.

> > > In such cases an XML Schema validator will correctly validate the _XML_
> > > input, but derive a unique post-Schema validation Infoset (psvi)
> according
> > > to which schema was used to validate. Are you are suggesting that ? the
> > > datatype model theory operate on the "psvi" graph rather than the input
> XML
> > > character stream?
> >
> > The behavior of RDF plus datatypes must be fully specified in the RDF plus
> > datatypes specification.  If it is possible to implement a significant
> > portion of the specification using an XML Schema validator, either because
> > the specification explicitly references XML Schema datatypes or in the
> > extraordinarily-unlikely event that the independently-specified RDF plus
> > datatypes specification happens to have a close semantic relationship to
> > XML Schema datatypes, then that is a (very) happy occurence.  I happen,
> > moreover, to believe that it should be the goal of the RDF Core Working
> > Group to build any RDF plus datatypes specification so that this is the
> > case.
> 
> I wholeheartedly agree. My point is to expose some of the possible seemingly
> insignificant inconsistencies so that any holes might be appropriately
> plugged. Working with types like "integer" is relatively simple because we
> all know one when we see one (it is a simple type). XML Schema provides a
> framework for defining an unlimited number of complex XML types, and until
> the issues regarding XML and RDF are better sorted out, datatypes in RDF,
> insofar as they include XML datatypes, is going to be frought with
> difficulties. Thankfully formalisms _do_ exist for defining XML lexical
> types (e.g. references in http://www.openhealth.org/RDF/RDFSyntaxFormal.html
> , XDuce http://xduce.sourceforge.net ). You can see that the pattern that
> describes the RDF XML 1.0 syntax itself is reasonably compact. (such
> formalisms do not trip up on the very basic <foo></foo> vs. <foo /> issues
> that trip up the psuedo-EBNF grammars).

That is fine.  Please feel free to point out any problems you see.

My stance is quite simple.

1/ RDF should incorporate XML Schema datatypes.

2/ The XQuery data model provides a relatively easy interface to get the
   resultant typed literals into RDF.

3/ If semantic typing is needed, it can be done using a slightly modified
   version of the code in an implementation of the XQuery data model.
   (Basically, you need direct access to the part of the implementation
   that performs the lexical-to-value mapping for XML Schema datatypes.)

> Jonathan

Peter F. Patel-Schneider
Bell Labs Research

Received on Tuesday, 4 December 2001 09:02:39 UTC