Re: namespaces

> 1. The essence of a "context" is a list of propositions
> which disambiguates a particular proposition.  I choose
> to talk about namespaces because a)they are lists of
> propositions which are already defined for RDF/OWL;
> b)the fundamental ones, like rdf and rdfs, are intended
> to be the foundation & definition for all other propositions.

There is your misunderstanding.

If you replaced the word "namespace" with "ontology", your statements  
would make sense.

<http://www.w3.org/2002/07/owl> is the URI for the OWL ontology*.  
Dereferencing it with the correct headers will retrieve a  
representation of the ontology in some RDF serialization format --  
you'll get back a sack full of triples.

Usually a namespace prefix is defined:

   owl = http://www.w3.org/2002/07/owl#

so that abbreviations can be used for OWL terms defined in that  
ontology -- e.g., owl:Class.

The namespace (prefix) is merely a syntactic element which makes it  
easier to write statements using the elements of the ontology. It is  
not itself a meaningful entity in the domain; just an artifact of  
serialization. The meaningful entities are the ontology and the  
classes and properties it describes.

If you read a little further along in the OWL guide, you'd see  
something like

   <owl:Ontology rdf:about="">
     <rdfs:label>My ontology</rdfs:label>
     ...
   </owl:Ontology>

-- i.e., the URI of the ontology denotes an instance of owl:Ontology.  
I believe that's the entity you're looking for.

Now, to decide whether an OWL ontology truly is the same as a  
'context' would require more detailed definition of the term... and  
I'm no logician, so I will not undertake that discussion. I hope you  
see, however, that 'namespace' is not the term for which you are  
looking.

-R

* Don't get too wrapped up in any other definition you have in mind  
for "ontology" -- in this context it approximately means an inter- 
related collection of statements about some classes and properties.  
Ontologies can import other ontologies.

Received on Saturday, 30 August 2008 04:48:47 UTC