W3C home > Mailing lists > Public > www-rdf-interest@w3.org > November 2000

Re: A triple is not unique.

From: Sergey Melnik <melnik@db.stanford.edu>
Date: Tue, 21 Nov 2000 12:31:23 -0800
Message-ID: <3A1ADB9B.D60AB90E@db.stanford.edu>
To: Seth Russell <seth@robustai.net>
CC: RDF Interest Group <www-rdf-interest@w3.org>
Seth Russell wrote:
> ...
> If we do it the way you prescribe above, then yes "A triple is unique".   But
> my stating now that [Bush wonThe Election] is certainly different from (not
> identical to) the Electoral College stating [Bush wonThe Election].  Certainly
> the context of utterance is what gives a triple its unique identity.

I definitely agree with you on that.

> I think I stand on firm ground when I say that.  [Word&Object]  So if a triple is to
> represent a unique entity ("A triple is unique"), then we would need to
> Skolemize the entire node, the nodes to which it arcs and the nodes that arc
> to it.
> 
> I disagree:  a triple is not unique.

A sufficiently smart application that crawls your Web site will not just
record

	[Bush wonThe Election]

The fact that this statement exists does not add anything to what we
already know.

Rather, the app will store something like that:

	[[Bush wonThe Election]  isSignedBy  <Seth's cryptographic signature>]
or
	[<Seth's URL>  contains  [Bush wonThe Election]]

For the above scenario, I don't see the need to Skolemize more than a
single statement.

Sergey
Received on Tuesday, 21 November 2000 15:13:29 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:46 GMT