W3C home > Mailing lists > Public > semantic-web@w3.org > July 2005

Usefulness of DirectObject

From: John McClure <jmcclure@hypergrove.com>
Date: Mon, 11 Jul 2005 12:26:23 -0700
To: <semantic-web@w3.org>
Message-ID: <MGEEIEEKKOMOLNHJAHMKEEFADOAA.jmcclure@hypergrove.com>

I thought I'd outline several arguments why DirectObject should be a metaclass
of fundamental importance to the Semantic Web. Recapping the approach in a
short-hand manner:

<rdf:Description rdf:ID='X'>
       <DirectObject rdf:ID='Y'>
	   <DirectObject rdf:ID='Z'/>

1. DirectObject implements the n-ary relation resource.
	a. DirectObject allows identity to be associated with a predicate.
	b. DirectObject allows properties to be associated with a predicate.
	c. DirectObject names resources that are both a Class and a Property.

2. DirectObject represents a basic Ontology Design Pattern.
	a. DirectObject appropriately represents a key grammatical concept.
	b. DirectObject functionally distinguishes nouns from qualified nouns.
	c. DirectObject features predicate-verbs with tenses and negatives.
	d. DirectObject is justified using an elementary data model.
	e. DirectObject leads to higher quality and easy-to-use ontologies.

3. DirectObject improves integration with other technologies.
	a. DirectObject directly supports UBL naming requirements.
	b. DirectObject names are wholly superior as XHTML2 property names.
	c. ECMA language support (via anonymous predicate-verbs) is automatic.
	d. Topics can now be about Class instances OR DirectObject instances.

I haven't discovered any downside for DirectObject, beyond its applicability
only within OWL-Full environments. If someone can identify an alternative to
DirectObject, I'd sure like to know how it would simultaneously provide for
integration with other technologies, in particular, with UBL, XHTML2, ECMA, and
Topics, for that is my ultimate goal.

John McClure
Received on Monday, 11 July 2005 19:26:16 UTC

This archive was generated by hypermail 2.4.0 : Tuesday, 5 July 2022 08:44:53 UTC