Proposed minutes for 6/26 from Christopher Welty on 2003-06-27 (www-webont-wg@w3.org from June 2003) (original) (raw)

Minutes of WebOnt Meeting 6/26/2003

Agenda: http://lists.w3.org/Archives/Public/www-webont-wg/2003Jun/0350.html IRC log: http://www.w3.org/2003/06/26-webont-irc

Next Meeting: 2003-07-03, Carroll to scribe.

====== Participants

Dan Connolly (Chair) Jerome Euzenat Mike Smith Evan Wallace Mike Dean Sandro Hawke (co-scribe) Jeremy Carroll Chris Welty (Scribe) Deborah McGuinness Tim Finin Peter Patel-Schnieder Charles White Jos DeRoo Herman TerHorst Jim Hendler Jeff Hefflin Guus Schreiber Sean Bechhofer Ian Horrocks

====== Regrets

====== Approval of Last Weeks Minutes

Approved

====== Action Review

===========COMPLETED

  1. "sameAs" error in OWL Language Reference 31 March 2003

Deemed adequate response. ACTION: Jim Hendler. Send response. ACTION: Jeremy Carroll. Generate test case.

  1. OWL S&AS comment - owl:imports

ACTION: Jeff Heflin - waiting for approval of response to send.

  1. Re: privacy in OWL

ACTION: Mike Smith. Guide to xref any Ref. change.

  1. OWL comment - blank nodes in OWL DL

Named data ranges Blank nodes used more than once. Jeremy Carroll B1, B2 proof in progress. Syntactic restrictions on unnamed indiv. (Covered) 17. OWL S&AS Comment - owl:imports

Proposed response sent to WG by Heflin. ACTION: Jim Hendler to review.

  1. Problem with PET of sections 7.1.5 and 7.1.6 of OWL test cases

WAITING FOR ACK. ACTION: Jim Hendler to confirm our response.

  1. inadequate reference in Owl REF

Guus had a link to S&AS built-in datatypes that are problematic. Now in the RDF semantics. S&AS points to that.

ACTION: Jeremey Carroll to send RDF link to Guus, Peter, Mike S.

  1. OWL and Published Subjects

ACTION: Jim to respond.

  1. QA Review of owl-semantics

Sandro sent proposed reply.

ACTION: Dan, Jim, Guus to review reply. ACTION: Guus to send proposed words re xrefs to all docs. ACTION: Peter, Guus, Frank, MKS All editors add link from their document to ALL others.

============CONTINUED

  1. Restriction, DeprecatedClass in OWL Language Reference 31 March 2003

ACTION: Guus Schreiber will craft process response.

  1. unspecified domain,range in App. C of OWL Language Reference 31 March 2003 CONTINUED: 12Jun ACTION: Mike Dean to respond.

  2. OWL Reference comment - RDF Schema for OWL

Guus S. sent propose response. Approved

ACTION: Guus S. to send response.

  1. Re: privacy in OWL

ACTION: Guus S. to propose response. (moved to JimH)

  1. daml:item

ACTION: Jim Hendler. Respond with pointer to Issue 5.5.

  1. RDFCore Comments on OWL Reference

Guus sent partial response. WAITING FOR ACK.

ACTION: Frank van Harmelen to respond to one open element.

  1. Comments on Owl REF document

DONE: 12Jun ACTION: Guus (sent to WG, approved by Dan)

ACTION: Guus to send.

  1. Non-global Keys

Compound keys not in response. ACTION: Guus to find record of decision and propose Issue to be postponed. ACTION: Jim Hendler to add to proposed response.

  1. OWL comment - owl:OntologyProperty

DONE: 12Jun ACTION: Guus. 12Jun ACTION: Jeremy. CONTINUED. 12Jun minute: Jeremy to respond ACTION: Guus to make Ref consistent with resolution

  1. Language Overview Document editorial comments

DONE: 12Jun ACTION: Frank. 12Jun minute: Frank to respond, editorial

WAITING FOR ACK. We think.

ACTION: Guus to find our response.

  1. Could owl:sameAs reference non-OWL resources?

12Jun ACTION: Pat. CONTINUED. 12Jun minute: Pat to respond

ACTION: Guus to check status.

  1. QA Review of owl-semantics

Sandro sent proposed reply.

ACTION: Jeremy, Jeff. All editors add link from their document to ALL others.

===================Agenda Item 3 - Sameas

http://lists.w3.org/Archives/Public/www-webont-wg/2003Jun/0345.html RESOLVED by consensus to accept Guus' proposal ABSTENTIONS: Jerome Euzenat ACTION: all editors to update

===================Agenda Item 4 - Compound Keys

proposal to open and postpone the issue Sentiment of group in favor - no vote

Jeremy summarizes - OWL lacks the ability to express keys of more than one attribute "It is very common in database systems to have primary keys composed of multiple fields, and in OWL, using InverseFunctionalProperties it is possible to do something similar with one field, but ..." technical work still in progress

Ian - may be possible - not necessarily a design flaw in OWL

ACTION Jim write up an issue description

===================Agenda Item 5.0 Dave Beckett's response

Beckett not satisfied with response of group, as per http://lists.w3.org/Archives/Public/public-webont-comments/2003Jun/0068.html

possibly send "what Sean wrote" (on RDF to OWL)

DanC wanted to add beckett to dissenters

jimh - are our syntax checkers sufficient for this?

jjc - tools do not conform to abstract syntax at this time

mikes - why translate from RDF triple to OWL abstract syntax?

danc - not a requirement (tried to say this reverse transformation is pretty closely related to the entailment testing task)

seanb - to do the reasoning youhave to understand the translation

danc - do we have implementors who have implemented the entire spec, as opposed to just satisfying the tests

danc - reverse mapping not required - just getting the "Right answers"

jeffh - semantics document uses abstract syntax, but documents talk about concrete syntax, and people need to do that mapping in their head

jimh - are people willing to evaluate Sean's document to make it (more) official

ACTION: JeffH report back to the WG on whether SeanBs document contributes to the issue

===================Agenda Item 6.0 Test LC Update

jjc - propose to reopen semantic layering. DL and Full consistency are not the same according to PFPS Tests that make that assumption maybe need to be rewritten

danc - document is different from WG decision

jos - we get an inconsistency

jjc- annotation property example at http://lists.w3.org/Archives/Public/www-webont-wg/2003Jun/0295 empty universe example at http://lists.w3.org/Archives/Public/www-webont-wg/2003Jun/0257

jim - is this just for annotation properties?

jjc - one is. The empty universe has to do with finite universes.

jim - As chair I reopen issue ACTION: (implied on MikeS) to update issues list with "semantic layering" reopened

[thanks to Sandro for scribing while ChrisW participates]

jjc - some work on test document required to fix owl:Thing owl:oneOf rdf:nil contradiction in owl full, but not in DL, because it can have the empty universe.

ChrisW - owl:Thing owl:oneOf rdf:nil . is a contradiction in OWL Full but not in OWL DL

pfps: it is in the syntax

jjc: or owl:Thing owl:EquivalentClass owl:Nothing jjc: and owl:Thing must have instances in owl full, since it has itself.

ianH: there are several nasty examples with finite universes

pfps: the whole idea of owl full was to pump up the universe, so you dont need to worry about typing.

jimh - just subtype OWL:Thing from RDF:Resource

pfps - not clear that would do it.

jimh - if we could describe this in one paragraph (as a caveat) then OK to close

jimh - need an owner for this issue. Volunteers?

...silence...

jim - I should have found an issue owner BEFORE re-opening. It's been so long.

pfps on the road for the next two weeks

pfps: I don't think I beleive the if-and-only-if statement

ianh: weaken it to "if" and we might be okay

peter's msg with the one-way entailment is at http://lists.w3.org/Archives/Public/www-webont-wg/2003Jun/0265

ianh: if it's entailed in DL it's entailed in Full, but not necessarily the other way around.

from Guide: every valid DL conclusion is a valid Full conclusion. PFPS: that's correct.

pfps: If it's owl full consistent, it's owl dl consistent.

ACTION: JJC summarize impact on test

ACTION: IanH notify FrankVH

ACTION: DanC to contact PatH and W3C colleagues

Next meeting: July 3rd, Hender chair, Carroll scribe.

================adjourned