W3C home > Mailing lists > Public > www-rdf-comments@w3.org > July to September 2003

Re: [closed] horst-01; was: RDF Semantics: RDFS entailment lemma

From: <herman.ter.horst@philips.com>
Date: Wed, 30 Jul 2003 20:16:56 +0200
To: pat hayes <phayes@ihmc.us>
Cc: www-rdf-comments@w3.org
Message-ID: <OFDAEC1F00.C224336A-ONC1256D73.00638138-C1256D73.00648B5F@diamond.philips.com>

Pat,

Thank you for your reaction 

http://lists.w3.org/Archives/Public/www-rdf-comments/2003JulSep/0012.html

to my comment.

I believe that it is a wise decision of the RDF Core Working Group to
change (weaken) the normative RDFS semantic conditions on subClassOf 
and subPropertyOf in order to preserve the completeness claim
of the RDFS entailment lemma.
The error that I noted in this lemma indeed vanishes by means of this
change.  As you note, there is now a direct correspondence between
many of the RDFS semantic conditions and many of the RFDS closure
rules.
It is interesting that the new version of the RDF Semantics document
presents, in addition, in Section 7.3.1, a number of entailment 
rules that are valid (sound but not complete) if the previous
iff semantics is considered instead.

Best regards,

Herman ter Horst

=====================================================================
Herman,

with reference to your comment
http://lists.w3.org/Archives/Public/www-rdf-comments/2003AprJun/0042.html

archived as horst-01.

First, many thanks for your close attention to the technical matters 
and your helpful comments, which have required the WG to re-examine 
several of its earlier decisions.

The WG has resolved
http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2003Jul/0025.html
to proceed as follows.

The RDS semantic conditions on subClassOf and subPropertyOf have been 
weakened to be 'only if' rather than 'iff' conditions; under these 
conditions, the entailments that you noted are no longer valid.

Details can be checked in the current editor's draft at 
http://www.ihmc.us/users/phayes/RDF_Semant_Edit_Weak.html#rdfs_interp 
The document mentions the stronger (iff) conditions under the heading 
'extensional semantic conditions'
http://www.ihmc.us/users/phayes/RDF_Semant_Edit_Weak.html#ExtensionalDomRang
, and describes some inference rules which are valid under these 
stronger conditions, including rules which cover the case you describe
http://www.ihmc.us/users/phayes/RDF_Semant_Edit_Weak.html#RDFSExtRules
; but it does not claim to offer a complete set of such rules for 
these stronger conditions.

A proof of the completeness of the (somewhat shortened) list of RDFS rules
http://www.ihmc.us/users/phayes/RDF_Semant_Edit_Weak.html#RDFSRules
  WRT the weakened semantic conditions is provided in appendix B
http://www.ihmc.us/users/phayes/RDF_Semant_Edit_Weak.html#prf
. As you will see, most of the complexity of the proof has to do with 
the treatment of literals; the RDFS rules correspondence is now 
straightforward since all the semantic conditions map directly into 
forward implications.

Please reply to this message, copying www-rdf-comments@w3.org, 
indicating whether this response addresses your comment adequately.

Sincerely

Pat Hayes



-- 
---------------------------------------------------------------------
IHMC             (850)434 8903 or (650)494 3973   home
40 South Alcaniz St.             (850)202 4416   office
Pensacola                                                (850)202 4440 fax
FL 32501                                                 (850)291 0667 
cell
phayes@ihmc.us       http://www.ihmc.us/users/phayes
Received on Wednesday, 30 July 2003 14:18:19 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 21 September 2012 14:16:32 GMT