RE: Response to JC5

Hi!

>-----Original Message-----
>From: public-owl-wg-request@w3.org [mailto:public-owl-wg-request@w3.org]
>On Behalf Of Ian Horrocks
>Sent: Friday, May 15, 2009 6:54 PM
>To: OWL 1.1
>Cc: Peter Patel-Schneider
>Subject: Response to JC5
>
>The current draft [1] looks good to me -- speak up now if you disagree.

No, please stop!

"""
Therefore the Working Group will not be making 
any change to OWL 2 in response to this comment.
"""

This is NOT what I have communicated to TQ! The change of the names, as we
have discussed yesterday, is crucial:

  owl:subject   --> owl:annotatedSource
  owl:predicate --> owl:annotatedProperty
  owl:object    --> owl:annotatedTarget

And possibly also some explaining text, probably in NF&R and/or Primer, that
the annotation vocabulary is meant to be used exclusively for annotation
purposes in the specifically defined way, if there isn't such text already.

Apart from this, in my discussion with TQ, I became aware that there might
really be an issue with getting back to RDF Reification for annotations, at
least a principle one. Topbraid Composer directly supports the reification
of arbitrary RDF statements, and mixing such unrestricted custom reification
with the very precisely defined annotation of axioms would possibly lead to
a mess. 

In particular, since RDF reification is intended to happens at triple level,
while axiom annotation, although also technically only reifying a "main
triple", targets the complete triple /set/ encoding a given axiom. So,
depending on the view, the same reification statement could have different
scopes, and this may well confuse tools. Also, different types of axioms are
annotated in a different way, only some of them in a reification-style. So
while RDF reification would on the one hand be used by users also for other
things than creating OWL annotations, on the other hand it wouldn't cover
axiom annotation completely. 

I still cannot precisely point to the bit which will definitely break the
system, so I have to think more deeply about this topic, but I believe that
there would be real hidden threads lurking around for some existing tools,
if we would use the RDF reification vocabulary for annotations.

Anyway, the draft needs to be redrafted before we can send it. It's
currently not in line with my discussion with TQ.

Michael

>Peter: if you don't hear anything to the contrary please send it off
>tomorrow.
>
>Thanks,
>Ian
>
>[1] http://www.w3.org/2007/OWL/wiki/LC2_Responses/JC5

--
Dipl.-Inform. Michael Schneider
Research Scientist, Dept. Information Process Engineering (IPE)
Tel  : +49-721-9654-726
Fax  : +49-721-9654-727
Email: michael.schneider@fzi.de
WWW  : http://www.fzi.de/michael.schneider
=======================================================================
FZI Forschungszentrum Informatik an der Universität Karlsruhe
Haid-und-Neu-Str. 10-14, D-76131 Karlsruhe
Tel.: +49-721-9654-0, Fax: +49-721-9654-959
Stiftung des bürgerlichen Rechts, Az 14-0563.1, RP Karlsruhe
Vorstand: Prof. Dr.-Ing. Rüdiger Dillmann, Dipl. Wi.-Ing. Michael Flor,
Prof. Dr. Dr. h.c. Wolffried Stucky, Prof. Dr. Rudi Studer
Vorsitzender des Kuratoriums: Ministerialdirigent Günther Leßnerkraus
=======================================================================