W3C home > Mailing lists > Public > www-webont-wg@w3.org > March 2002

Re: LANG: A proposal for the layering problem

From: Dan Connolly <connolly@w3.org>
Date: 28 Mar 2002 15:59:54 -0600
To: Jeff Heflin <heflin@cse.lehigh.edu>
Cc: WebOnt <www-webont-wg@w3.org>
Message-Id: <1017352795.12257.50.camel@dirk>
On Thu, 2002-03-28 at 14:42, Jeff Heflin wrote:
> Dan,
> 
> I don't believe your description about partial understanding, i.e., that
> "the common syntax of RDF allows agents of various capabilities to
> extract the same set of facts from a document." If the WebOnt language
> (formerly known as OWL) provides additional semantics beyond RDFS, then
> an RDFS agent cannot expect to extract the same facts from a WebOnt
> document as a WebOnt agent would.

Er.. no?

An RDF parser will give you all the facts from any RDF file,
whether it uses DAML+OIL vocabulary or not.

> Since WebOnt has additional semantics,
> the WebOnt agent ought to be able to infer additional facts.

I guess I wasn't clear...

I'm not talking about entailed conclusions; just stated facts.

> The fact that an RDF Schema agent can determine that daml:disjointWith
> is a property or daml:UnambiguousProperty is class doesn't seem very
> useful to me. If this is "partial understanding," then what's the point?

Hard to explain briefly. Did you read the...

> The Extensible Languages document you cite says

ah; evidently so. But it didn't make the point about
partial understanding. Hmm...


> "The resource defining a
> namespace may be generic and allow definitions of the namespace in
> varying present or future languages." Note that it doesn't say "in a
> single language."

Well, other parts of it do talk about evolution within a language
as well as from one language to another.

> RDFS Schema is one language for providing definitions
> of a namespace.

To me, RDF Schema is a vocabulary of terms (e.g. subClassOf) to be used
within RDF, i.e. within the Resource Description Framework,
for describing resources such as properties and classes.

WebOnt should be another vocabulary of terms (e.g. disjointWith)
that can be used in the same description framework.

> All I am proposing is that we provide definitions of the
> namespace in a language other than RDF triples.

I wish I could make the point about how much we lose if we do that.

Perhaps by way of history:

1997: Channel Definition Format(CDF): its own vocabulary, its own
  syntax for stating ground facts. An early adopter
  of XML.

  OSD (software descriptions) and DRP (replication) follow soon after.
  Each has is written in XML (yeah!) but each has its own
  way of stating ground facts (booh!).

Sep '97: W3C push workshop
  http://www.w3.org/Architecture/9709_Workshop/
  where W3C lays the ground for an integrated
  metadata framework.

1999: Dublin Core integrates their vocabulary into
 RDF, establishing, e.g.
   http://purl.org/dc/elements/1.1/title
 as a name for its title property

2000: RSS 1.0: a vocabulary similar in expressiveness
  to CDF, but integrated into RDF.

2000: I can use RSS and Dublin Core vocabulary together
  to write a summary of what's up at W3C in one go:
  http://www.w3.org/2000/08/w3c-synd/#


Meanwhile, P3P is developed as a way of stating privacy
policies. That WG decided not to use RDF. Not fatal
to the semantic web, since I can develop
  http://www.w3.org/2000/10/swap/p3p/grokPolicy.xsl
which maps P3P syntax to the RDF model, but
I can't mix P3P vocabulary and dublin core
(and RSS and ...) in the same file to develop
integrated descriptions of the resources
I manage.

Early in the development of XML Schema, there
was a proposal to use RDF to describe
element types and so on...
  http://www.w3.org/TR/1998/NOTE-dcd-19980731

But the XML Schema WG decided not to integrate
with RDF. Again, not fatal to the Semantic Web:
their component model is a directed labelled graph, so 
the syntactic transformation is almost trivial.
They gave URI-names for the primitive
datatypes and each of the facets; that's key.
I haven't written grokXMLSchema.xsl to
convert to RDF yet. The dang XML Schema
language is so big that it looks too
arduous.

But I can't mix dublin core, RSS, P3P, and
XML Schema together to say what I want
to say.

Same story for XML Infoset...

Yes, we can always do post-hoc integration,
but it's so much harder than just using
a common framework in the first place.

First of all, when we do the post-hoc
analysis, we find bugs. My post-hoc
analysis of the P3P spec turned
up various questions that the P3P WG
doesn't have an answer to.
XML Schema interoperability is
not what I would have hoped.

Second, all this translation is
a huge drag on the network effect
of the Semantic Web.


> This language can
> coexist with RDF Schema and translation services can be written to allow
> maximal understanding between RDFS and WebOnt agents.

Yes, that's entirely possible. But it's not very interesting.

I'm pretty much obliged to help this WG thru the process
of building a W3C Recommendation based on whatever
technical consensus it comes up with. But I can
tell you right now that if it's not integrated
into the Resource Description Framework, my
heart won't be in it: it'll simply be
yet another legacy data format that I need
to write grokOntology.xsl converters for.

I could get into the development of an RDF2...
i.e. something where stating n-ary facts
isn't so painful... something with universal
as well as existential quantification...
that sort of thing.
But yet another XML syntax that can't be
automatcially merged/conjoined with data
written in other vocabularies? Yawn. Sigh. Trundle trundle...

> 
> Jeff
> 
> Dan Connolly wrote:
> > 
> > On Tue, 2002-03-26 at 10:22, Jeff Heflin wrote:
> > > Dan,
> > >
> > > Can you describe what you don't like about my proposal?
> > [...]
> > 
> > > Let me try to refute the likely arguments in favor of building on top of
> > > RDF schema:
> > 
> > None of those is the critical one.
> > 
> > The ciritical one is: partial understanding, as explained in [extlang].
> > Briefly: the common syntax of RDF allows agents of various
> > capabilities to extract the same set of facts from a document.
> > 
> > If, in order to introduce properties and classes
> > (such as disjointWith or UnambiguousProperty)
> > we have to change our syntax for stating facts, we lose this.
> > 
> > [extlang]
> >  Web Architecture: Extensible Languages
> >   W3C Note 10 Feb 1998
> > This Version:
> >   http://www.w3.org/TR/1998/NOTE-webarch-extlang-19980210
> > Latest Version:
> >   http://www.w3.org/TR/NOTE-webarch-extlang
> > Authors:
> >     Tim Berners-Lee <timbl@w3.org> W3C
> >   Dan Connolly <connolly@w3.org> W3C
> > 
> > esp the aircraft purchase order scenario
> > http://www.w3.org/TR/1998/NOTE-webarch-extlang-19980210#Scenario
> > 
> > --
> > Dan Connolly, W3C http://www.w3.org/People/Connolly/
-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/
Received on Thursday, 28 March 2002 16:59:49 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:57:48 GMT