[
Lists Home |
Date Index |
Thread Index
]
Thanks for the clarifications, and I agree these technologies
need to coexist. It would seem one could start from the
XML DTD/Schema types, get the local systems running, then
come in later and add RDF/OWL type descriptions.
What are the benefits of dual systems of classification?
len
From: jborden@attbi.com [mailto:jborden@attbi.com]
Len, I agree with what you say below.
To be clear, RDF is not _itself_ an ontology language.
But one can develop an ontology language e.g. DAML or
OWL which is based upon RDF, i.e. is written in RDF
syntax (triples) and at the same time XML (as a dialect
of RDF/XML).
At the same time, via XML datatypes, one can speak about
(e.g. make assertions about, or classifications of)
pieces of structured data in OWL.
In RDF you can assign a type (i.e. place into a
classification) with an "rdf:type" arc e.g.
<http://example.org/foo> rdf:type owl:Class .
now assuming that you have an XML Schema type you might
integrate this with OWL as such:
my:date-format rdf:type owl:DatatypeProperty .
so the idea that a fragment of XML conforms to a certain
XML Schema _type_ is the same as saying that such a
fragment belongs to a corresponding OWL _Class_.
So in summary, you might assign free-form
RDF 'individuals' to classes, and you might also assign
fragments of markup to classes, both using OWL (which is
the successor to DAML).
Both RDF and XML approaches to
classifications/ontologies need to coexist on the
semantic web.
|
