Re: Terminology Question concerning Web Architecture and Linked Data

On 7/25/07, Dan Brickley <danbri@danbri.org> wrote:
> >> 5. they are suitable for use in the same
> >>     circumstances.
> >>
> >> 6. they have dictionary definitions which are
> >>     word-for-word identical.
> >
> > No?  I would expect both of those would hold.  Why wouldn't they?
>
> In RDF, usage suitability can be a property of a name. For example, I
> try not to use names for RDF classes and properties that begin http://
> and whose domain name does not seem to be paid up for at least 2-3 years.

Interesting point.  But if they're not substitutable, then I wouldn't
call them aliases.  I suppose this relates to my earlier point that I
don't expect to see owl:sameAs used cross-domain.

It seems to me as if the folks who told Chris he could use owl:sameAs
were doing so using *indirect* identification, i.e. saying the
equivalent of "10 Downing Street" owl:sameAs "British government"[1]
(using my argument that introducing agency turns direct into indirect
identification).  I interpreted its definition as being in terms of
direct identification, but I suppose indirect isn't an unreasonable
interpretation.  But it does mean that the current definition is
either wrong or ambiguous.

 [1] http://www.w3.org/TR/webarch/#indirect-identification

Mark.
-- 
Mark Baker.  Ottawa, Ontario, CANADA.         http://www.markbaker.ca
Coactus; Web-inspired integration strategies  http://www.coactus.com

Received on Wednesday, 25 July 2007 14:55:28 UTC