W3C

SemWeb Deployment WG

01 Jul 2008

Agenda

See also: IRC log

Attendees

Present
Elisa Kendall, Jon Phipps, Ralph Swick, Danial Rubin, Diego Berrueta, Ed SUmmers, Alistair Miles, Guus Schreiber, Sean Bechhofer, Clay Redding, Daniel Maycock, Antoine Isaac, Ben Adida, Tom Baker
Regrets
Simone Onofri, Margherita Sini, Quentin Ruel
Chair
Guus
Scribe
Jon

Contents


<Ralph> Previous: 2008-06-24 http://www.w3.org/2008/06/24-swd-minutes.html

ADMIN

Guus: This is the last scheduled telecon

PROPOSED to accept minutes of the last telecon: http://www.w3.org/2008/06/24-swd-minutes.html

RESOLUTION: to accept minutes of the last telecon

RDFa

Ralph: Nothing we particularly need to talk about

<Ralph> meeting record: 2008-06-26 RDFa telecon

Ralph: On schedule for August proposed rec

Guss: We need to schedule a meeting for about that time.

<scribe> ACTION: [PENDING] Ben to prepare draft implementation report for RDFa (with assistance from Michael) [recorded in http://www.w3.org/2008/01/22-swd-minutes.html#action14]

Recipes

Guus: what's the progress on getting the Note published?

Jon: I have updated the Status paragraph but didn't get a chance to send the notification
... Ralph should look at it
... Diego found some additional errors in the example document, which I'll fix right after the meeting

<berrueta>

<scribe> ACTION: Jon and Ralph to publish Recipes as Working Group Note [recorded in http://www.w3.org/2008/06/10-swd-minutes.html#action03] in http://www.w3.org/2008/06/10-swd-minutes.html#action03] [CONTINUES]

<scribe> ACTION: [DONE] Diego to propose minimum RDFa metadata set for WG deliverables. [recorded in http://www.w3.org/2008/06/24-swd-minutes.html#action16]

[recorded in http://www.w3.org/2008/06/24-swd-minutes.html#action16]

Diego: see -> http://lists.w3.org/Archives/Public/public-swd-wg/2008Jun/0104.html "adding metadata with RDFa to W3C TR" [Diego 2008-06-29]

<scribe> ACTION: [PENDING] Ralph/Diego to work on Wordnet implementation of Recipes [recorded in http://www.w3.org/2008/01/22-swd-minutes.html#action20]

Vocabulary Management

Guus: realistic timescale?

Elisa: almost done, just need to validate
... should have by next week

Guus: we may start reviewing between telecons, but will have to see how that works out

Elisa: several people have found the doc to be valuable
... we were going to include some recommendation about the SKOS namespaces
... but will figure that out once we've started reviewing

SKOS

Guus: Antoine sent email on ISSUE-84

<Ralph> Proposal to postpone ISSUE-84 ConstructionOfSystematicDisplaysFromGroupings [Antoine 2008-07-01]

Antoine: considering that issue-84 is too complex to deal with in the time available
... issue-84 is borderline wrt SKOS application and I propose to postpone

PROPOSED: postpone ISSUE-84, reason given in message 0001 of July 2008

<Ralph> +1

RESOLUTION: postpone ISSUE-84, reason given in message 0001 of July 2008

Ralph: I'll update the issue list right now, no action needed

Guus: looking at ISSUE-86

SeanB: action on me and Alistair to compose some text, Alistair has seen
... suggestion is to follow practices from CoolUris and include in Appendix
... proposed resolution is to make no requirements but recommend authors should follow the recipes and CoolUris

Guus: ISSUE-72, ISSUE-73, ISSUE-75

aliman: just sent a mail suggestion some positions for each
... for issue-72, we make no statement
... for issue-75 suggest that we don't assert any property chains for exact match

<Ralph> exactMatch issues: ISSUE-72 ISSUE-73 ISSUE-75 [Alistair 2008-06-24]

aliman: issue-73, when we say related, we're saying there's an associative relationship, and from that perspective it's worth stating that they're disjoint

<aliman> http://lists.w3.org/Archives/Public/public-swd-wg/2008Jun/0097.html -> suggestions for exactMatch issues

Antoine: from the pint that we've already asserted semantics for matching properties
... I was afraid that Alistair's position was a step backward
... I don't have a strong objection but am uncomfortable

<Ralph> current specification of skos:exactMatch

Antoine: I support exactMatch disjoiint with related but not broadertransitive

aliman: i could live with not saying that exactmatch is not disjoint with any other property, but that users should check

edsu: no opinion about this

aliman: one of the difficulties is that we don't have any obvious use cases

Guus: that means we should follow the least commitment strategy

aliman: that means that we should say nothing formally on any of these issues

Antoine: would really like to make exactmatch transitive

Daniel: What are the arguments against saying transitive

<Ralph> We currently say "[skos:exactMatch] is typically used to indicate that two concepts are sufficiently similar that they can be used interchangeably in an information retrieval application."

aliman: making no statement allows people to draw their own conclusions
... if we _say_ that its transistive then we specify that you're drawing conclusions across mappings

Daniel: it would seem that you would want that to be transitive

aliman: I've never looked at the data, so that was my reluctance

Daniel: if there was another semantics for exactMatch then we should have another property

Ralph: I agree

edsu: but of course this may map across a number of concepts
... if it's transitive then there's bound to be drift

Daniel: exactMatch has a specific semantics that would seem to require transitivity

SeanB: if you make the explicit statement that they're transitive, then you have the possibility of rendering errors
... given the "sufficiently similar" wording

<Ralph> aliman: exactMatch is more for a specific application to use rather than a general statement

Daniel: then we need a different property
... exactmatch implies exact

Ralph: why don't we have a different property "similarMatch"

aliman: I can see both points of view

Guus: I can see Daniel's point that it needs to be transitive

SeanB: but "sufficiently similar" isn't exact enough

aliman: there may be assertions across mappings that requires careful checking of data

Daniel: exactmatch need to be exact

aliman: but this isn't an exact world
... these shouldn't ever be used in concept schemes

Ralph why not similarMatch

Daniel: How about nearlyExactMatch

<Ralph> Tom: "closeMatch"

Ralph: closematch

<Antoine> +1 with not changing the name

Guus: unless we have strong reasons I'd rather not change the name

many variations bandied about

Ralph: ok with exactMatch as long as there's an addition that exact == close to

Guus: this is why we're not using owl:sameAs

aliman: this is an issue of quality of exactness of match

Guus: propose to not change the name but add wording

<Ralph> PROPOSE: keep the name "exactMatch" but add a sentence saying that "exact" in this context means "sufficiently similar to" and not "identical to".

aliman: transitivity is just one entailment

<Ralph> PROPOSE: keep the name "exactMatch" but add a sentence saying that "exact" in this context means "sufficiently similar to" and not "identical to" and this relation is not transitive.

SeanB: seems like there's an inconsistency when you say woolily similar, but then say it's transitive, then you introduce the opportunity to compound errors

Guus: reluctant to change the name because it's already been deployed

<ed> Ralph++

Guus: who would be in favor of "closeMatch"?

aliman: these things should be so similar that you can swap em

<Ralph> I prefer "closeMatch" but would not object to keeping the name with the fuller explanation

SeanB: do you have to qualify the map

Guus: change the wording of exactmatch to say that it is sufficiently close and not transitive

Ralph: why would you feel that a transitive exact is better than owl:sameAs

<seanb> ++1 for what Antoine is saying

<aliman> ++1

Antoine: owl:sameAs comes with additional formal semantics that don't apply here

SeanB: what we're trying to represent here is application behavior, and very different from sameAs

Guus: straw poll

<Ralph> I don't feel a need for _both_ transitive exactMatch and also closeMatch

Guus: exactMatch is transitive

<Ralph> -1 to both transitive exactMatch and close

<seanb> This appeals to me as a solution, but I'm not a system developer :-)

Guus: introduce closeMatch as subproperty of exactMatch that is not transitive

Ralph: not sure if there's a use case to have both

Guus: typically exactmatch would be 1 to 1

aliman: we have no use cases for mapping across vocabularies
... not sure if it's a lack of use case or lack of data
... I can live without exactMatch

Daniel: why can't we have both
... wouldn't this represent a good compromise

aliman: if we keep both then closeMatch can't be a subproperty

<aliman> i was wrong, exactmatch could be a sub-prop of closeMatch

Alistair agrees with SeanB that this isn't necessarily so

<Ralph> I can live with both transitive exactMatch and closeMatch

Daniel: I can live with the last proposal of 2 properties, whether one is a subproperty or not

all agree with 2 properties

PROPOSED: ISSUE-72 is resolved by 1) adding to the skos data model a property "closeMatch" which is not transitive. 2) add to skos data model that exactMatch is transitive

<aliman> PROPOSED: ISSUE-72 is resolved by 1) adding to the skos data model a property "closeMatch" which is not transitive. 2) add to skos data model that exactMatch is transitive

<Ralph> +1

seconded Daniel

RESOLUTION: ISSUE-72 is resolved by 1) adding to the skos data model a property "closeMatch" which is not transitive. 2) add to skos data model that exactMatch is transitive

Guus: reference editors please add wording for this
... leave it to them to figure out subproperty relationship
... but first want to have it right in the reference

<scribe> ACTION: Alistair and Sean to propose text to implement the resolution of issue-72 [recorded in http://www.w3.org/2008/07/01-swd-minutes.html#action05]

Guus: issue-73...

<Ralph> ISSUE-73 ExactMatchDisjoints

aliman: think this changes now and we can take a stronger position

PROPOSED: issue-73 is resolved by skos:exactMatch is disjoint with skos:broaderTransitive and skos:related

Alistair seconds

RESOLUTION: issue-73 is resolved by skos:exactMatch is disjoint with skos:broaderTransitive and skos:related

Guus: last issue, issue-75
... property chain axioms

<Ralph> ExactMatchInclusions

SeanB: my inclination is to not do this, but could go either way

Guus: don't see any need to define this here
... I'm happy with the proposal that for the moment there are no property chain axioms

Antoine: I could support this

Guus: Close this issue by asserting that there are no property chain axioms until there is evidence to support such axioms
... would be useful to include the rationale

PROPOSED: Close Issue-75 by asserting that there are no property chain axioms until there is evidence to support them

<Ralph> PROPOSED: Close ISSUE-75 by asserting that there are no property chain axioms as there is no evidence yet to support them

Antoine seconds

RESOLUTION: Close ISSUE-75 by asserting that there are no property chain axioms as there is no evidence yet to support them

<Ralph> ISSUE-86

<seanb>

SeanB: We haven't yet closed ISSUE-86

sean reads text of email

Ralph: "makes no requirement" is not as strong as "strongly suggests"

seanb: happy to strongly suggest

Ralph: I'd prefer "does not require but strongly recommends"

<ed> Ralph++

<Ralph> PROPOSE: Close ISSUE-86 with and Appendix saying "URIs are used to identity resources of type skos:Concept and skos:ConceptScheme. The SKOS Reference does not require specific behaviour when dereferencing those URIs. It is, however, strongly recommended that publishers of vocabularies follow the guidelines for Best Practice Recipes [REF] and Cool URIS [REF]."

sean seconds

RESOLUTION: Close ISSUE-86 with and Appendix saying "URIs are used to identity resources of type skos:Concept and skos:ConceptScheme. The SKOS Reference does not require specific behaviour when dereferencing those URIs. It is, however, strongly recommended that publishers of vocabularies follow the guidelines for Best Practice Recipes [REF] and Cool URIS [REF]."

Guus: planning of telecon: 22July and another a week later
... 22 July for SKOS candidate recommendation, the other for RDFa

<seanb> I am definitely not here on the 22nd July

seanb: not available 22 July

aliman: one more week would be better

<Ralph> [I'm at risk during August]

Guus: like to have reviewers no, version available for review bu August
... happy to review reference
... chairs will look at this and be intouch
... editors please start implementing the changes

<scribe> ACTION: [PENDING] Ed to investigate what text could be added to primer re. concept co-ordination [recorded in http://www.w3.org/2008/04/22-swd-minutes.html#action02]

<scribe> ACTION: [PENDING] Guus to write primer text re: broaderGeneric and equivalence w/r/t subclass [recorded in http://www.w3.org/2008/03/18-swd-minutes.html#action08]

<scribe> ACTION: [PENDING] Alistair to check the old namespace wrt dereferencing [recorded in http://www.w3.org/2008/05/06-swd-minutes.html#action03]

<scribe> ACTION: [PENDING] Antoine and Ed to add content to Primer about irreflexivity [recorded in http://www.w3.org/2008/05/06-swd-minutes.html#action06]

<scribe> ACTION: [PENDING] Alistar to update the history page adding direct link to latest version of rdf triple [recorded in http://www.w3.org/2008/06/17-swd-minutes.html#action01]

<scribe> ACTION: [DONE] Editors of the Use Cases to clean up the lists of requirements in light of resolutions [recorded in http://www.w3.org/2008/05/07-swd-minutes.html#action02]

see http://lists.w3.org/Archives/Public/public-swd-wg/2008Jun/0093.html

<scribe> ACTION: [DONE] Antoine to propose that we postpone ISSUE 84. [recorded in http://www.w3.org/2008/06/24-swd-minutes.html#action10]

<scribe> ACTION: [PENDING] SKOS Reference Editors to specifically flag features at risk for Last Call. [recorded in http://www.w3.org/2008/06/24-swd-minutes.html#action17]

<scribe> ACTION: [PENDING] Sean to draft response to comment about namespace. [recorded in http://www.w3.org/2008/06/24-swd-minutes.html#action12]

<scribe> ACTION: [PENDING] Sean to post comment to OWL WG re annotation requirements. [recorded in http://www.w3.org/2008/06/24-swd-minutes.html#action06]

<scribe> ACTION: [PENDING] SKOS Reference Editors to propose a recommended minimum URI dereference behaviour [recorded in http://www.w3.org/2008/06/24-swd-minutes.html#action11]

ADJOURNED

Summary of Action Items

[NEW] ACTION: Alistair and Sean to propose text to implement the resolution of issue-72 [recorded in http://www.w3.org/2008/07/01-swd-minutes.html#action05]
 
[PENDING] ACTION: Alistair to check the old namespace wrt dereferencing [recorded in http://www.w3.org/2008/05/06-swd-minutes.html#action03]
[PENDING] ACTION: Alistar to update the history page adding direct link to latest version of rdf triple [recorded in http://www.w3.org/2008/06/17-swd-minutes.html#action01]
[PENDING] ACTION: Antoine and Ed to add content to Primer about irreflexivity [recorded in http://www.w3.org/2008/05/06-swd-minutes.html#action06]
[PENDING] ACTION: Ben to prepare draft implementation report for RDFa (with assistance from Michael) [recorded in http://www.w3.org/2008/01/22-swd-minutes.html#action14]
[PENDING] ACTION: Ed to investigate what text could be added to primer re. concept co-ordination [recorded in http://www.w3.org/2008/04/22-swd-minutes.html#action02]
[PENDING] ACTION: Guus to write primer text re: broaderGeneric and equivalence w/r/t subclass [recorded in http://www.w3.org/2008/03/18-swd-minutes.html#action08]
[PENDING] ACTION: Jon and Ralph to publish Recipes as Working Group Note [recorded in http://www.w3.org/2008/06/10-swd-minutes.html#action03] in http://www.w3.org/2008/06/10-swd-minutes.html#action03]
[PENDING] ACTION: Ralph/Diego to work on Wordnet implementation of Recipes [recorded in http://www.w3.org/2008/01/22-swd-minutes.html#action20]
[PENDING] ACTION: Sean to draft response to comment about namespace. [recorded in http://www.w3.org/2008/06/24-swd-minutes.html#action12]
[PENDING] ACTION: Sean to post comment to OWL WG re annotation requirements. [recorded in http://www.w3.org/2008/06/24-swd-minutes.html#action06]
[PENDING] ACTION: SKOS Reference Editors to propose a recommended minimum URI dereference behaviour [recorded in http://www.w3.org/2008/06/24-swd-minutes.html#action11]
[PENDING] ACTION: SKOS Reference Editors to specifically flag features at risk for Last Call. [recorded in http://www.w3.org/2008/06/24-swd-minutes.html#action17]
 
[DONE] ACTION: Antoine to propose that we postpone ISSUE 84. [recorded in http://www.w3.org/2008/06/24-swd-minutes.html#action10]
[DONE] ACTION: Diego to propose minimum RDFa metadata set for WG deliverables. [recorded in http://www.w3.org/2008/06/24-swd-minutes.html#action16]
[DONE] ACTION: Editors of the Use Cases to clean up the lists of requirements in light of resolutions [recorded in http://www.w3.org/2008/05/07-swd-minutes.html#action02]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.133 (CVS log)
$Date: 2008/07/08 02:03:55 $