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

Re: new semantics initiative

From: Patrick Stickler <patrick.stickler@nokia.com>
Date: Sun, 16 Jun 2002 15:12:08 +0300
To: "ext R.V.Guha" <guha@guha.com>, Pat Hayes <phayes@ai.uwf.edu>
CC: Brian McBride <bwm@hplb.hpl.hp.com>, RDF Core <w3c-rdfcore-wg@w3.org>
Message-ID: <B9325B48.16DB2%patrick.stickler@nokia.com>

On 2002-06-15 2:26, "ext R.V.Guha" <guha@guha.com> wrote:

> 
> This is not just a single namespace. It is a space of namespaces. So,
> some terms from owl  will be in it. As will some terms from daml-s and
> so on. Any term inroduced by any future language to deal with logical
> machinery (e.g., log:implies by cwm/euler), should be designatable as
> going into this namspace so that RDF doesn't think its a simple triple.
> 
> guha

I like to use the term 'functional vocabulary'.

A vocabulary is a set of terms with associated semantics.
Some applications may use/require only some of those terms, and
may require also a few terms from other vocabularies. Thus we
can define a aggregate, functional vocabulary which is the set
of terms that are relevant to a particular function or set
of related functions, where the semantics are not defined by
that functional vocabulary but by the source vocabularies from
which they are taken.

Thus, PRISM could be seen as a functional vocabulary. Likewise,
this "space of namespaces" Guha refers to is a functional vocabulary.

Whether a given vocabulary is defined using a single
XML namespace is irrelevant insofar as the concepts of vocabulary
and functional vocabulary are concerned. Yes, it is often
editorally convenient to ground all terms of a vocabulary in
the same namespace, if XML is being used, but merely convenient,
not required.

I think that alot of the "debate" that has occurred recently both
on the wg list and on rdf-interest rises from the desire of some
folks to make what is simply convenient (namely a 1:1 correspondence
between RDF vocabulary and XML Namespace) an actual requirement,
so that that convience is guarunteed.

I do not consider such a constraint to be in the best interests
of the semantic web -- and it has already been demonstrated that
such a 1:1 correspondence does not hold for key vocabularies
already in use.

I think we need to, once and for all, make clear that the
naming model used by XML is *not* the naming model used by
RDF and the concept of an XML namespace has no direct
equivalent in RDF and any percieved equivalence between an
RDF (functional) vocabulary and an XML namespace is just an
illusion arising from the fact that RDF uses XML for its
serialization and thus must at some point define namespaces,
and often, the URIref denoting the vocabulary is used as
the XML namespace prefix for those terms.

Cheers,

Patrick

> patrick hayes wrote:
> 
>> 
>>> At 12:07 12/06/2002 -0500, patrick hayes wrote:
>>> 
>>> [...]
>>> 
>>>> What is wrong with URI inspection?
>>> 
>>> 
>>> Questions:
>>> 
>>>   o what uri prefix should be used?  Is it ok to insist on an http:
>>> prefix?
>> 
>> 
>> I would guess so. I would expect that it would be done the same way
>> that the W3C handles the RDF and RDFS vocabulary, by a URL linking to
>> a set-in-stone page.
>> 
>>>   o how will names in this namespace be allocated?
>> 
>> 
>> Do you mean how procedurally? Thats up the W3C. I would guess that a
>> WG would submit some kind of application to some internal secretariat,
>> or something like that. Isnt this kind of stuff all set out in the W3C
>> process manual somewhere? For example, we are proposing to create an
>> rdfd: namespace, right? Like that.
>> 
>> Pat
>> 
>> 
> 
> 
> 

--
               
Patrick Stickler              Phone: +358 50 483 9453
Senior Research Scientist     Fax:   +358 7180 35409
Nokia Research Center         Email: patrick.stickler@nokia.com
Received on Sunday, 16 June 2002 08:07:53 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:49:19 EDT