RE: HTTP URIs for real world objects

Just to clarify a couple of points:
- when I refer to the use/mention issue it is in the sense of understanding that www.example.com/psi/apple can be "used" as being the thing pointed to (that URI might dereference, for example, to a string of data that, for the owner of the domain, is to be used as being the thing named); or it can be "mentioned" as something that is a marker or token, standing as a proxy for the thing, elsewhere in addressable network space or not; My issue today is that there is no mechanism in the RDF standard to tell me if you intend it stand for the formal or the latter case. Topic Map PSId's do.
- In this standard, there are two concepts that should not be confused, although used often interchangeably: Published Subject Identifier (PSId) is an arbitrary URI that is used to attribute unequivocal identity to some subject of discourse: if two different "topics" - the subjects' proxies in networked space (with different names, properties, etc) - nonetheless have the same PSId, a conformant processor must merge the topics as they are considered to be representations of the same subject. A Published Subject Indicator (PSI) is, IMO, something of a misnomer and prone to confusion but, is an optional addressable information resource that has been put up by someone to give humans an opportunity to disambiguate otherwise imprecise, ambiguous or simply opaque Topic names or PSIds...
You could think of PSIds as doing the machine processing part of identity, whereas PSIs are the human face. Indeed, I would support the slogan of Topic Maps being the "semantic web with a human face" precisely because it has these formal distinctions. It is not that I maintain that RDF is less useful or powerful - it just does something else, and does it well.

Wikipedia entry UIRs are indeed probably the best example of PSIs - that include in many cases even disambiguation pages too - but they are *not* PSIds... not that Wikipedia always gets it "right", but that's not the point: they are someone's assertion and - contrary to what some may believe is my assertion - I do *not* support some central authority responsible for issuing PSIds, just a recognition of the conceptual separation on the one hand between use and mention of a URI; and on the other between an arbitrary identifier and some (set of) resources that provide human disambiguation.

Peter
PSId= www.justbrown.net*, just to disambiguate this popular name ;-)

* you will notice that there is a difference between the PSId (that I intend to maintain) and the actual PSIs loaded as a resource (which might change over time)...

-----Original Message-----
From: Danny Ayers [mailto:danny.ayers@gmail.com] 
Sent: 17 January 2008 13:23
To: KANZAKI Masahide
Cc: martin.hepp@uibk.ac.at; Peter F Brown; Peter Ansell; Bernard Vatant; Reto Bachmann-Gmür; Leo Sauermann; public-sweo-ig@w3.org; semantic-web@w3.org
Subject: Re: HTTP URIs for real world objects

I just searched for "wikipedia rdf", and the the first entry on the
results page suggests Google has rather an eccentric view of this PSI:
[[
Resource Description Framework - Wikipedia, the free encyclopedia
The title of this resource, which is published by Wikipedia, is 'Tony
Benn'. However, RDF puts the information in the formal way that a
machine can ...
en.wikipedia.org/wiki/Resource_Description_Framework - 67k -
]]

-- 

http://dannyayers.com

Received on Thursday, 17 January 2008 12:50:21 UTC