See also: IRC log, previous 2008-11-04
RESOLVED to accept minutes of the late telecon http://www.w3.org/2008/11/04-swd-minutes.html
ACTION: Ben review RDFa Use Cases and propose transition to Group Note [recorded in http://www.w3.org/2008/09/30-swd-minutes.html#action02] [CONTINUES]
benadida: we're continuing on a bi-weekly basis
-- life after rec
... the folks at drupal the cms, have prepared a timeline for rdfa in
drupal
http://groups.drupal.org/node/16597
TomB: you have proposed some resolutions to remaining issues?
<Ralph> [Recipes] proposed resolution for remaining issues
Diego: should we go through them one by one?
TomB: i don't think so, unless there is discussion
Ralph: i concur with all 4 proposals
TomB: would anyone like to discuss?
RESOLVED to postpone issues 24, 30 and 98 and close 60 as per http://lists.w3.org/Archives/Public/public-swd-wg/2008Nov/0003.html
Diego: might be helpful to get other people in the working group looking at it, not sure if the timing is right ... would like to discuss the document at some point
TomB: i agree we would need to assign reviewers
to move this towards note status, but right now we have our hands full w/
skos
... lets move on with skos for now, and come back to it in a few weeks
seanb: is it right we can't add RDFa to REC documents?
Ralph: that is currently the state, pubrules don't allow it, i can revisit that
seanb: aliman and i discussed this, i figure it
wouldn't take long to put this in our SKOS Reference, and i think it would
send the right message
... would be willing to fold it in
Ralph: would be wonderful
TomB: Ralph could you check on the rdfa usage in the pubrules? is that within the scope of this working group?
Ralph: i can take an action for that
seanb: i tried to do this with my docs, and i had html entities which caused some problems with the rdfa dtd
Diego: is this for existing html entities? I haven't seen it
Ralph: i remember danbri saying he used numeric entities ...
http://lists.w3.org/Archives/Public/semantic-web/2008Nov/0151.html
ACTION: Ralph to report on use of RDFa metadata in Recommendations. [recorded in http://www.w3.org/2008/11/18-swd-minutes.html#action02]
ACTION: Guus to look at OWL documents for review [recorded in http://www.w3.org/2008/10/21-swd-minutes.html#action10] [CONTINUES]
<Ralph> [for Sean; the message from DanBri that mentioned using numeric entity rather than is http://lists.w3.org/Archives/Public/public-rdfa/2008Nov/0004.html ]
TomB: lets start with the actions, and go back to discussion
ACTION: Guus and Jeremy to give concrete implementation examples of the use of rdfs:label w/ SKOS [recorded in http://www.w3.org/2008/10/07-swd-minutes.html#action10] [CONTINUES]
ACTION: Guus to propose answer for issue 186 [recorded in http://www.w3.org/2008/10/21-swd-minutes.html#action02] [DONE]
<Ralph> ISSUE 186 - draft response [Guus]
TomB: ok lets start with ISSUE-135
<Ralph> issue 135; rdfs:label
seanb: this is concerneing the subproperty
relationship with rdfs:label
... and whether pushing out of owl DL is a good idea
... we already have things outside of owl DL so this isn't the issue
... one way of tackling this would be to assert that they are annotation
properties
... might be easier to migrate to owl2
... i think of the labling properties as annotation properties, i'm not clear
if this would constitute a substantial change, would be interested in what
alistair and others have to say
Guus: rdfs:label is currently an annotation property?
<Ralph> OWL Annnotations
seanb: pretty sure
Guus: i can't see a real reason against it
Antoine: would it have consequences with what we say about the range of the property?
<Ralph> "The sets of object properties, datatype properties, annotation properties and ontology properties must be mutually disjoint. Thus, in OWL DL dc:creator cannot be at the same time a datatype property and an annotation property." -- http://www.w3.org/TR/owl-ref/#Annotations
<Ralph> "The object of an annotation property must be either a data literal, a URI reference, or an individual." -- http://www.w3.org/TR/owl-ref/#Annotations
seanb: i believe that one can specify ranges of annotation properties in owl2
Guus: it only makes sense if we can specify value restrictions, cardinality and sub-properties
seanb: as i understood it we would be able to range/domain and sub-properties -- not sure about cardinality
Guus: the non-owl user will ignore this anyway
Ralph: i think it's pretty useful to have subproperty of relationship there, i think it doesn't make sense to have it any other way
seanb: i imagine most applications will be using sub-property anyway to get the behavior that they want
aliman: i don't know what's happening w/ owl2 --- just heard bits and pieces about annotations
seanb: i'm hearing that this is a potential solution to this issue
Guus: i support it
Ralph: +1
aliman: abstain
<aliman> In http://www.w3.org/TR/owl2-syntax/#Annotations I see nothing about annotation property axioms...
Guus: your question is then 'does this change our design' ... i consider it a small refinement
<aliman> specifically .. http://www.w3.org/TR/2008/WD-owl2-syntax-20081008/
Ralph: any implementation that was conformant
is still conformant
... we've done due diligence to adding this to our issues list
seanb: we would be removing the assertion that it's a datatype property, and adding the new assertion
Guus: we can just say this was an error, and correct the error
seanb: are you happy with that alistair?
aliman: i don't know
TomB: if it's a small refinement that's ok -- but could it be arguedthat this is a substantial change?
seanb: i'm uncomfortable with labeling it as an error ... it seemed like a more appropriate way of typing the property
TomB: if i can ask simple question, why is this
not an rdf:property?
... an alternative would be just to remove the datatype assertion
aliman: i never had a strong preference one way or the other ... but others do rely on it
Guus: if owl people can add the triple we are
fine
... if we remove the owl:datatype statement we are fine
... a less commmitting resolution
<Ralph> skos:*Label Class & Property Definitions
seanb: but why don't we do that with *everything* ?
<aliman> From SKOS Reference:
<aliman> """
<aliman> We can, therefore, use OWL to construct a data model for representing thesauri or classification schemes "as-is". This is exactly what SKOS does. Taking this approach, the "concepts" of a thesaurus or classification scheme are modeled as individuals in the SKOS data model, and the informal descriptions about and links between those "concepts" as given by the thesaurus or classification...
<aliman> ...scheme are modeled as facts about those individuals, never as class or property axioms. Note that these "facts" are facts about the thesaurus or classification scheme itself, such as "concept X has preferred label 'Y' and is part of thesaurus Z;
<aliman> """
Ralph: seems we only used this with notations
aliman: early on we made a decision that skos would be an owl full ontology
Guus: maybe we should separate the issues? i don't think use of annotation properties would change the design
seanb: it does open the can of worms: should perhaps other properties in skos be annotation properties
<Ralph> [our RDF does in fact only explicitly state <owl:DatatypeProperty rdf:about="http://www.w3.org/2008/05/skos#altLabel" />
aliman: if you are dealing with individuals in a KOS you don't even need annotation properties ... the only use cases where you need annotations are when you start taking bits and pieces of skos and using them elsewhere
TomB: maybe we can take a decision on the next call, I would rather we not rush into this ... get a proposed resolution up on the list
Guus: are there other cases where skos properties where they are subproperties of owl annotation properties?
seanb: no
TomB: it would be good to have this proposal in
writing, and to make clear it doesn't change conformance
... that we can consider in the next call
Ralph: we have declared everything in reference to owl, and not rdf -- so it requires owl reasoning ...
seanb: well it requires knowledge of the relationshiops to the owl schema
Guus: minimal amount of owl reasoning
... it would perfectly fine to add the rdf triples, can only be a gain
Ralph: if you have RDFS reasoning and have the OWL schema loaded you'll be in good shape -- would be good enough
<aliman> fine with me to add p rdf:type rdf:Property assertion to schema for all property p in SKOS vocabulary
ACTION: Sean to propose a resolution to ISSUE-135 [recorded in http://www.w3.org/2008/11/18-swd-minutes.html#action06]
<Ralph> +1 to meeting next week to close issues
ACTION: Sean to add rdf:type and rdf:Property assertions to the skos schema [recorded in http://www.w3.org/2008/11/18-swd-minutes.html#action07]
<aliman> +1 to meet next week
RESOLVED to meet on November 25th
seanb: issue-147
-> http://lists.w3.org/Archives/Public/public-swd-wg/2008Nov/0064.html
aliman, Antoine, Guus : support
RESOLVED close issue #147 per http://lists.w3.org/Archives/Public/public-swd-wg/2008Nov/0064.html
<Ralph> issue 147; Notations as plain literals
seanb: can anyone look at the current version of the reference where i stuck in some text as an appendix about the namespace change issue
seanb: it's the latest working version