W3C home > Mailing lists > Public > www-webont-wg@w3.org > September 2003

Minutes 4Sep

From: Jeremy Carroll <jjc@hplb.hpl.hp.com>
Date: Fri, 05 Sep 2003 10:53:40 +0100
Message-ID: <3F585D24.3080303@hplb.hpl.hp.com>
To: Dan Connolly <connolly@w3.org>
Cc: www-webont-wg@w3.org

Executive Summary:
  no decisions
  routine business
  discussion most animated for item 9 concerning Full nonentailments

Agenda:
http://lists.w3.org/Archives/Public/www-webont-wg/2003Sep/0033
IRC:
http://www.w3.org/2003/09/04-webont-irc
(Dan/Sandro please fix permissions)


> --- 1. convene, take roll, assign scribe, review record
> scribe: Jeremy
> 


Present:
DanC, Jeff_Heflin, Sandro,
Evan_Wallace, JeremyC, JeromeE, HermanT,
Jos, ChrisW


regrets: Hendler, Schreiber, Hori, McGuinness, Dean,
   Gibbins, Patel-Schneider 

    Mike Smith, Obrst,  Sabbouh, White


> 
> PROPOSED: to accept
> 
>   MINUTES Teleconference Web Ontology Working Group - August 21 20 03 
>   Smith, Michael K (Thursday, 21 August)
>   http://lists.w3.org/Archives/Public/www-webont-wg/2003Aug/0096.html
> 
> as a true record.


So carried.


> --- 2. Review agenda and misc actions, plan next meeting
> 
> consider outreach telcons
> 
> PROPOSED: to resume weekly schedule, starting
> 	Thu 11 Sep at noon Boston time


So agreed, agenda at discretion of chairs (i.e. no decision about outreach 
telecons)

Scribe still needed.


Continued ACTIONS: 
> 
> ACTION: Jim will report back on status of Gene Ontology Consortium re
> OWL.




> --- 3. Comment handling
> 


Jeremy asked about the change log in the editors' draft.
DanC advised that the change log should reflect changes against the CR version.


> 
> --- 4. DL Syntax, B1/B2
> 

Completed action:

> ACTION: Peter Patel-Schneider: to look over Jeremy's B1 B2 proof
> revision. Continued until next week.
> 

Continued action:

> ACTION: Dan C. - Add links to implementation report describing Guus'
> summary from editors meeting and Jeremy email.
> 


The proof attempt was:
http://lists.w3.org/Archives/Public/www-webont-wg/2003Jun/0017
as modified by:
http://lists.w3.org/Archives/Public/www-webont-wg/2003Jun/0294

The refuting example was:
http://lists.w3.org/Archives/Public/www-webont-wg/2003Aug/0142
and thread.


Jeremy reported that:
- Peter's action as complete
- Peter had found a significant flaw with the proof.
- Jeremy still believes the result is correct.
- Peter believes the result to be incorrect.
- Neither belief is proven.
(note Peter was absent with regrets)

Jeremy offered to take an action to attempt to fix the proof (but noted 
that the fix and review would take time). No action was assigned, and 
further discussion was postponed to item 10.

> --- 5. RDF Internationalization
> 


Completed actions:


> ACTION: Jeremy to send his discussion of some of the issues re
> xml:lang and literals to WG.
> 
> ACTION: Guus S. will review.
> 
> ACTION: Guus Schreiber will send some examples of use of xml:lang to
> webont mailing list.


RDF Core WD will be published 5th September, and will specifically ask for 
comments on this issue. Informally Jeremy indicated that a consensus 
comment from WebOnt would be useful.

A brief discussion as to whether Guus' msg:
http://lists.w3.org/Archives/Public/www-webont-wg/2003Sep/0027

could constitute such a comment, but there was insufficient informed debate.
Jeremy noted that specifically discussing the wrapper solutions in the 
comment would be helpful to RDF Core.




> --- 6. Test review
>

Completed action:

 
> ACTION: Jeremy C. to study DL 909 and report back.



Jeremy indicated that a new version of the editors draft with the fix 
(10billion to 1billion) would be available next week.

Jos has changed some proposed test by the character #
(as reported:
http://lists.w3.org/Archives/Public/www-webont-wg/2003Aug/0113
)

> 
> --- 7. Exit criterion: two complete OWL Lite consistency checkers
> http://www.w3.org/2001/sw/WebOnt/impls#exit
> 

Continued actions:


> ACTION: Jim Hendler - Report PELLET status re complete OWL Lite
> consistency checkers
> 
> ACTION: Ian Horrocks - Report Cerebra and Racer status re complete OWL Lite
> consistency checkers
> 






> --- 8. Exit criterion: Each test (except the extra credit tests) is demonstrated
> 	to be passed by some implementation
> 


Continued action:


> ACTION: Charles White will collect data on test detail. Send him lists
> of test you have passed.
> 




> 
> --- 9. Exit Criterion: two reasoners implementing useful subsets of OWL Full ...
> 


Completed actions:


> ACTION: Sandro - Report reasoning status re useful subsets of OWL
> Full.
> 
> ACTION: Joss de Roo - Sandro will ask him to report reasoning status
> of Euler.
> 
> ACTION: Jeremy - Will ask Dave Reynolds re reasoner status over useful
> subsets of OWL Full.
> 



Extended dicussion of pass/fail/incomplete for non-entailment and 
consistency tests.

Some classes of reasoners, particularly those by OWL Full implementors 
cannot prove nonentailments, but merely fail to prove entailments in these 
cases. Sandro wished these to be reported as incomplete (but see after 
telecon e-mail traffic)

There was discussion as to whether our exit criteria should be changed, 
DanC suggested that it would be cheaper to make all these tests extra 
credit (well at least *proving* all these tests as extra credit)

Discussion continues.



> 
> -- 10. exit criterion: two owl syntax checkers passing all tests
>


Completed actions:

 
> ACTION: Sean B. - Report officially on species validation syntactic
> checks.
> 
> ACTION: Peter Patel-Schneider - Will report on DL once problem with
> Galex fixed (exp. in a week).
> 
> 


Continued actions:

> 
> ACTION: Ian Horrocks - Confirm that Network Inference passes all owl
> syntax checks.


 > ACTION: Jim Hendler - Check with Bijan re owl syntax checkers passing
 > all tests
 >



It was noted that Sean has not yet fully implemented the feature at risk, 
whereas Peter has.
e.g.
http://www.w3.org/2002/03owlt/disjointWith/consistent004

Scribe note the discussion drifted, without a formal change of agenda item.
Topic: names for tests.

Sandro and DanC noted that the syntax tests were one two or more per test 
case depending on the number of files.
It was also noted that the semantic tests are often two per test case (one 
using the DL semantics and one using the Full semantics)

DanC argued that it would be helpful if we had each individual test 
individually named. There was significant resistance from the test editors 
against duplicating tests to give new names.

ACTION jeremy To think about test naming and report back.


> 
> 
> --- 11. ontologies in the web: MIME types, file extensions, etc.


Agendum dropped, and meeting adjourned.
Received on Friday, 5 September 2003 07:16:32 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:02 GMT