RE: Summary (I mean it this time): URIs : How to find the ontologies behind them

> -----Original Message-----
> From: ext Peter F. Patel-Schneider 
> [mailto:pfps@research.bell-labs.com]
> Sent: 10 April, 2003 16:38
> To: Eli@SemanticWorld.Org
> Cc: www-rdf-interest@w3.org
> Subject: Re: Summary (I mean it this time): URIs : How to find the
> ontologies behind them
> 
> 
> 
> From: "Eli Israel" <Eli@SemanticWorld.Org>
> Subject: Summary (I mean it this time): URIs : How to find 
> the ontologies behind them
> Date: Thu, 10 Apr 2003 14:50:07 +0300
> 
> > 
> > Ok, I think I am ready to summarize this thread.
> > 
> > 
> > It is a given that the URIs in an RDF/OWL document serve 
> only as unique
> > identifiers.  One can not infer meaning from the content or 
> stucture of the
> > URIs.  Similarly, the URI  need not be a valid web address, 
> and if it is, it
> > need not point to anything having to do with the ontology.  
> Likewise,  the
> > web authority has no obligation to present, in any way, 
> information about
> > the URIs that refer to it.
> 
> Not all web addresses, and not even all http: addresses, have 
> any chance of
> having any ``authority''.

That's your opinion. Others, including myself, think differently.

It is true that determination ownership and authority will likely
ultimately be a legal process, but I believe that if the SW
is ever going to be of use to commercial and other entities 
where issues of ownership, authority, trust, and accountability
are concerned, the issue of who controls the denotation of a
given URI will need to be resolved, and the issue of ownership
and authority for specifying such denotation addressed by
the SW community.

> > That being said, we agree that it would be a good idea for 
> the web authority
> > refered to in the URI of a resource to provide information 
> about that
> > resource, either at the address pointed to by the URI, or 
> through another
> > mechanism.
> 
> Provided that there is single such an organization, yes, this 
> is a good
> idea.  In fact, even in the absence of an authority, it is a 
> good idea to
> have information concerning a URI reference accessible in 
> some standard
> fashion.

We all seem to agree on this point. That given a URI, one
should be able to obtain a description of the resource denoted
in a consistent, ideally standardized, manner.

> > However, even if an authority provides information about a 
> resource, it can
> > not be seen as either authoritative or complete.  
> Information about the
> > resource garnered from other sources should be given equal 
> weight.  The
> > provision of information by the web authority is, then, a technical
> > convenience, and nothing more.
> 
> This I agree with totally, except that in some circumstances, 
> one might
> want to make this information ``first among equals''.

Isn't the phrase "first among equals" an oxymoron?

If one member of a set is considered to take precidence over
the others, then it is not equal to the others.

I understand the (IMO tongue in cheek) meaning of the phrase,
but I don't consider oxymorons (however amusing and clever)
to be valid as a basis for SW architecture.

In any case, I think it should be completely up to the consumer
of knowledge to decide how to valuate input, and whether to
treat authoritative knowledge as "first among equals" or in 
any other manner. The key criteria for this, of course, is that
there is a clear concept of authoritative knowledge and a means
of differentiating authoritative knowledge from non-authoritative
knowledge. Using the existing web architecture and the well defined
machinery of http: URI web authorities seems one reliable way to do
this, and is the approach I am exploring.

I think in practice, we will see that authoritative knowledge obtained
via a web authority (for URIs with web authorities) will usually be
given special value over all other sources of knowledge. Time will tell.

Cheers,

Patrick

--
Patrick Stickler, Nokia/Finland, (+358 40) 801 9690, patrick.stickler@nokia.com

Received on Friday, 11 April 2003 01:33:53 UTC