Re: Meta-classes? (layered architecture)

Ian Horrocks wrote:
> 
> We need to be clear as to whether or not it is intended that DAML-ONT
> classes (relations) denote classes of objects (binary relations
> between objects) or whether they could also denote classes of classes,
> classes of relations etc. (it doesn't matter what we consider an
> "object" to be - it is simply something that our ontology language
> treats as an atom).

It is intended (by this contributor) that DAML-ONT
classes could also denote classes of classes,
classes of relations (e.g. TransitiveProperty) etc.

This has been called the "anyone can say anything
about anything" principal... that there's just
one sort of assertion in RDF. That's probably
not a sufficient motivation/justification,
but it's what I have at my disposal just now...

> In OIL, it was explicitly decided that, in the base language, the
> denotation of classes and relations should be sets of objects and
> binary relations between objects, and that meta-extensions, if
> required, would be provided in additional language layers (yet to be
> defined). The rationale behind this layered architecture is that we
> should extend RDFS in relatively small increments (in the same way
> that RDFS is a small increment w.r.t. RDF).

In this sense, I don't view RDFS as an extension to RDF;
that is: a document that uses terms from the RDFS
vocabulary is still an RDF document. RDFS is,
in this sense, more of an application of RDF than
an extension of it... just like XHTML is an application
of XML, not an extension of it.


> This makes it more likely
> that widespread agreement can be reached on the specification of any
> given increment, allows for extension along multiple axes, and gives
> applications designers a choice about the language layer that they
> want to commit to, rather than the "all or nothing" choice that is
> presented by a single large language.
> 
> It has been suggested that DAML languages will also have a layered
> architecture, and that further layers (DAML-RULES, DAML-LOGIC) will
> follow. If this is the case then I think it makes sense to stick to
> the simple thing for now and leave meta-classes for DAML-META.

That would mean that DAML-META is some other language,
rather than just more terms added to the vocabulary
in an existing language. That's very different from
the way I expect DAML, among other RDF applications,
to evolve.

I expect RDF tools to have partial understanding
of all documents, whether they can exploit
the semantics of any particular bit of vocabulary or not.

cf

Web Architecture: Extensible Languages 
W3C Note 10 Feb 1998 
     http://www.w3.org/TR/1998/NOTE-webarch-extlang-19980210 

-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/

Received on Thursday, 19 October 2000 00:44:55 UTC