W3C home > Mailing lists > Public > public-swbp-wg@w3.org > January 2005

RE: [SE] Composite Identification Schemes on the Semantic Web

From: Bernard Vatant <bernard.vatant@mondeca.com>
Date: Sat, 22 Jan 2005 13:06:24 +0100
To: "Phil Tetlow" <philip.tetlow@uk.ibm.com>
Cc: <public-swbp-wg@w3.org>, <tom.croucher@sunderland.ac.uk>
Message-ID: <GOEIKOOAMJONEFCANOKCCEFLFHAA.bernard.vatant@mondeca.com>


Phil, and all

I'm well aware of the SE Task Force, and in fact discovered the paper of Tom listed at
http://dl-web.man.ac.uk/~panz/swse/.

Actually I was making my mind about putting this identity/identification issue on SWBP WG
table, but you beat me at it ...

As written in my private mail to Tom Croucher and al, and since you were kind enough to
push it in the public space :)) some thoughts about how this question could be tackled in
the SW Activity framework, and beyond. I am not completely sure if this is only a Semantic
Web issue, although it is critical to the Semantic Web deployment. SW Activity seems to be
interested mostly (only?) by identification of resources by URIs, although Tom's paper
among others clearly shows the limits of such an approach.
The identity issue is certainly in the scope of SE Task Force. Of course identification of
resources/subjects/topics across applications is an integration issue in a Semantic
Software environment, but this is not IMO the whole of identity issue. It's also in the
scope of VM, for example, since identifying identical concepts across distributed
vocabularies is also a core issue. As I pointed out a few days ago, there are today about
400 different URIs in the Swoogle Ontology Dictionary somehow defining "Person", there is
no reason why this figure will not keep up growing, a situation I tend to call the
"Semantic Mess" [1].

So far I was mainly brainstorming about it in my blog, trying to figure the various tracks
available. We've been working for quite a while in the OASIS Published Subjects TC, since
2001, to figure out the best recommendations for Subject Indicators and Subject
Identifiers, and basically we have piled up open issues, and there again, it's all about
URI-based identity. Heuristic approaches based on various kind of "Subject Identity
Measure" [2] or what Tom calls "Composite Identification" are indeed to consider, but they
somehow question the URI-based mechanism as the final solution for identification.

In other words, I think time is ripe to shift from the notion of absolute identity based
on static identifiers (names, URIs) to the notion of identification protocols, providing
"context specific identification" [3].

[1] http://universimmedia.blogspot.com/2005/01/how-many-person-concepts-in-semantic.html
[2] http://universimmedia.blogspot.com/2004_08_01_universimmedia_archive.html
[3] http://osiris.sund.ac.uk/~cs0tco/eswc2005.pdf

**********************************************************************************

Bernard Vatant
Senior Consultant
Knowledge Engineering
bernard.vatant@mondeca.com

"Making Sense of Content" :  http://www.mondeca.com
"Everything is a Subject" :  http://universimmedia.blogspot.com

**********************************************************************************

> -----Message d'origine-----
> De : Phil Tetlow [mailto:philip.tetlow@uk.ibm.com]
> Envoye : samedi 22 janvier 2005 10:44
> A : Bernard Vatant
> Cc : public-swbp-wg@w3.org; tom.croucher@sunderland.ac.uk
> Objet : [SE] Composite Identification Schemes on the Semantic Web
>
>
>
>
>
>
> Bernard
>
> I currently help coordinate the W3C's task force on the Application of the
> Semantic Web in Software Engineering and am a friend of Tom Croucher. As
> such Tom have been kindly sharing his thoughts with me on composite
> identification schemes for the Semantic Web and has passed on your mail
> below. As such I am sure that members of the task force would be most
> interested to read the material you have listed. We also have a public
> mailing list, so if you would like to contribute to our cause, friendly and
> constructive debate is always warmly welcomed.
>
> Kind Regards
>
> Phil Tetlow
> Senior Consultant
> IBM Business Consulting Services
> Mobile. (+44) 7740 923328
> ----- Forwarded by Phil Tetlow/UK/IBM on 22/01/2005 04:30 -----
>
>              Tom Croucher
>              <tom.croucher@sun
>              derland.ac.uk>                                             To
>                                        Phil Tetlow/UK/IBM@IBMGB
>              21/01/2005 12:21                                           cc
>
>                                                                    Subject
>                                        Fwd: Your papers about Identity on
>                                        the Web
>
>
>
>
>
>
>
>
>
>
>
>
> ------- Forwarded message -------
> From: "Bernard Vatant" <bernard.vatant@mondeca.com>
> To: tom.croucher@sunderland.ac.uk, j.r.c.geldart@durham.ac.uk,
> rguha@us.ibm.com
> Subject: Your papers about Identity on the Web
> Date: Fri, 21 Jan 2005 14:28:46 +0100
>
>
> Bonjour
>
> I've read with most interest your respective papers on identity
> http://osiris.sund.ac.uk/~cs0tco/eswc2005.pdf
> http://tap.stanford.edu/CoIdent.pdf
>
> I'm trying to gather contributions to those identity issues, that I
> consider also as the
> most critical for the future of semantic technologies, and even IT at
> large. I've set for
> a few months a blog about it at http://universimmedia.blogspot.com to
> gather relevant
> papers, relexions and debates. You are of course welcome to jump in, post
> comments, or new
> topics if you feel like it (I will send you a formal invitation to do so
> in this latter
> case).
>
> I've been thinking for a while to go beyond this informal effort, and try
> to gather a
> community towards some standardisation effort in this domain, but am not
> sure what would
> be the best framework for it. In the W3C Semantic Web Activity, I'm afraid
>
> the debate
> would be at risk to be too much focused on URI-based identity. I've also
> been
> participating in the OASIS Published Subjects group, but there again the
> focus has been on
> a very specific URL-based identification protocol ...
>
> Any thoughts welcome
>
> Best regards
>
>
> **********************************************************************************
>
>
> Bernard Vatant
> Senior Consultant
> Knowledge Engineering
> bernard.vatant@mondeca.com
>
> "Making Sense of Content" :  http://www.mondeca.com
> "Everything is a Subject" :  http://universimmedia.blogspot.com
>
> **********************************************************************************
>
>
>
>
> --
> Tom Croucher
>
Received on Saturday, 22 January 2005 12:46:50 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:09:41 UTC