W3C home > Mailing lists > Public > www-webont-wg@w3.org > May 2002

Dark triples are not inherently nonmonotonic

From: Jonathan Borden <jonathan@openhealth.org>
Date: Thu, 30 May 2002 19:17:25 -0400
Message-ID: <036401c20830$2eab0ed0$0a2e249b@nemc.org>
To: "WebOnt WG" <www-webont-wg@w3.org>
Cc: "pat hayes" <phayes@ai.uwf.edu>, <guha@guha.com>

Pat, did you reall write:
http://lists.w3.org/Archives/Public/www-webont-wg/2002May/att-0302/00-part ?

Now I agree that the mechanism you proposed to indicate darkness e.g.

owl:List a owl:Dark

is problematic from a monoticity perspective, but to suggest that

 but the exact details do not matter to the point being made here.

and that _dark triples themselves are inherently nonmonotonic_ is going a
bit to far, eh?

For example than triples, are rather than dark, let's consider: colored
triples, and so a triple becomes a quad:

subject predicate object color .

where color is a URIref just like the others.

Now assume that for all current RDF documents, the parser generates a color
rdf:White, which corresponds to an asserted triple in the MT.

Now suppose that the color of the triples (really quads but go with this),
is a function of some attribute on the <rdf:RDF> root element of any RDF
document e.g.

<rdf:RDF color="red">
    <rdf:Description rdf:ID="foo">
        <ex:shape rdf:resource="#square">

which generates the following N-Quad:

<#foo> ex:shape <#square> rdf:red .

of course the color doesn't need to be rdf:red, but could be any URI, but
that really doesn't matter.

So what could possibly be nonmonotonic about that?

Now assume that dark triples are triples with the color rdf:black.

Are you really suggesting that triples are monotonic, but quads are
_inherently nonmonotonic_ ? Surely not.

So, as with everything, the exact details do matter to the point being made
here, and to suggest that dark triples are inherently nonmonotonic can only
be true if your specific definition of what a dark triple is and how it is
created causes the nonmonotonicity.

Received on Thursday, 30 May 2002 19:22:51 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 23:04:30 UTC