Semantic Web Best Practices & Deployment Working Group
18 Nov 2004

Adminstrivia

review of minutes of oct 28 - minutes accepted

review of minutes from f2f - minutes accepted

telecon times, in response to email from Chris

guus: he has a point

ralph: is this week slip a temporary aberation?

david: we talked at f2f about this [temporary adjustment, or permanent adjustment]

ralph: I propose we look ahead into January to decide when we should meet

david: thanksgiving, christmas and new year don't work; keep to slipped schedule for the rest of this year

ACTION: david to propose a telecon schedule for the new year

Ralph notes that a meeting on jan 6, the bi-weekly schedule would fit well with the tech plenary

ralph: we are agreed to meet on 02 Dec 2004, regrets from me as there is a w3c meeting

RESOLVED next meeting is 02 Dec 2004

Action Review

ACTION BenB read ODM documents
-- withdrawn

ACTION gary ng review ODM
-- DONE

ACTION Ralph to ask WG for feedback on requirement to embed RDF/XML markup in an XHTML document
-- DONE

ACTION danbri circulate links for his existing feedback/review to dawg
-- DONE

ACTION philT look at gary ng's message, see what actions if any this wg should take
-- DONE

ACTION chrisw approach sophia about units and measures particpating
-- DONE

ACTION guus to note on numeric ranges after the xml datatypes TF has finished
-- continued

ACTION libby to make that note into a document to read for the f2f by 25th oct
-- DONE

ACTION guus send jeremy pointer about numeric ranges and XMLS
-- continued

ACTION jjc to send around pointers on HTML TF
-- DONE

ACTION libby to send pointers to list in preparation for f2f
-- DONE

ACTION Alistair make explicit in skos core doc the fact that you're trying to deal with potential for multiple thesauri using the ame terms, overlap etc., different from paper publishing world
-- DONE

ACTION Ben to send this statement [regarding RDF/A] to HTML WG via email
-- DONE

ACTION Brian and DanBri need to talk about what need to do for Wordnet document to be good enough
-- continued

ACTION David to reword the statement on RDF A to HTML WG
-- done

ACTION JJC review SPARQL WD re http://www.w3.org/TR/2004/WD-rdf-sparql-query-20041012/#extendedtests
-- DONE

ACTION Phil to write up concerns about RDF/A on email
-- DONE

ACTION Steve to email on concerns for RDF in XHTML
-- DONE

ACTION VM TF to compile list of sample vocabs for the note
-- continued

ACTION VM TF with help from Guus to find thesaurus like example and high end ontologies to section 3
-- continued

ACTION David to contact Eric Miller re his interest in joining the RDFTM TF
-- continued

ACTION find someone to do the review the part of UML about TM

david: whose ACTION?

Steve: asked me to ask Lars Marius to do that. I asked Lars Marius and he went pale

david: mark action complete

ACTION Jeremy Clarify which parts of UML docs HP is most interested in reviewing
-- continued

ACTION Steve to finish rdftm TF description
-- DONE

TECH PLENARY / SWBPD MARCH FTF

guus: there was a generic request from the tag whether other working groups want to meet with the tag during the Technical Plenary

david: do we want to meet with the tag? we could state our position on xhtml working group

ralph: not sure that is an appropriate topic, but the so called httpRange-14 issue is. httpRange-14 ought to be one of our issues too

steve: is relevevant to rdftm task force

<aliman> +1 on talking to TAG on HTTP range

ralph: will hit other tf's too e.g. vocab management; should talk to the tag about that

david: that issue is important to us - we've done it one way but not sure its the right way; meeting with the tag would a good idea for us

pepper: would be good idea to have a general discussion

Alistair: this is the biggest issue for the porting tf

phil: lets discuss on what topics we need to discuss on the list

ralph: we could do that, but the timing is somewhat tight
... I suggest the range14 dicussion may be better with the whole tag

ACTION: Ralph take up TAG-SWBP agenda CG tomorrow

ralph: I said we want to meet two of those days, prefering thu/fri
... also asked if it were practical to meet for 4 days

steve: would like to do tech work on rdftm - could take two days

guus: we could have breakout groups on two days

RDFTM Task FORCE PROPOSAL

david: I'm happy with the TF description. any objections?

david: we have a lot of interest in the rdftm task force; want to get it underway. any objections?

ralph: are we quorate for those who agreed to participate in the task force? we have three TF members on the call, so thats ok

david: steve do you have commitments from the folks listed that they want to participate

steve: yes

ralph: can they make the telcon time

steve: yes

david: steve will you take an action to get them at the next telecon

guus: I'll help

steve: thanks guus

ACTION: guus to introduce new members from rdftm task force to the wg

ralph: propose approve task force creation

RESOLVED: RDFTM creation approved

steve: description of work explains what we do
... we need to start note on existing practice
... we need to collect test cases to evaluate proposals
... snippets of both rdf and tm for translation back and forth

david: can you talk to folks like nikita and danc
... they have specific concerns - valuable to collect in use case

steve: iso group met in dc
... informed them about the tf
... general reaction was extremely positive
... some discussion of brining more folks in
... membership of w3c can be an issue
... what you may see is an initiative at a higher level to establish a liason group between jtc1 to allow formal input
... have a recognised position for reviewing

<pepper> WG3 resolution: "WG3 expresses its support for the W3C's initiative in setting up a task force to address the issue of RDF/Topic Maps interoperability and encourages the active participation of members of the Topic Maps community."

david: we can always send a document draft to iso for comment

steve: that would probably satisfy them

david: could record in tf description

steve: could add that

guus: what you have written is good enough
... we are required to request feedback from the dependent parties
... you have to reach consensus if they give comments

ralph: identifying them in particular as a group makes it explicit we hope to hear from them

david: can we proceed in the face of an objection from iso?

ralph: we'd handle it like any other public comment

guus: we'd have to explain why he should override to the director

ralph: there would be a lot of procedural things involved in setting up a formal liason structure
... task force could be "well underway" before it could be set up
... but we could look at it if there is a strong need
... it is not completely impossible

steve: there is other work in iso that overlaps w3c work

ralph: there are a number of liason things going on between w3c and JTC 1 and this could be added
... I would propose the TF proceeds without that and they can come back

steve: they will; the chair has an action
... contact will be made and there can be a general discussion

ralph: if you could provide a url for 13250 that would be great

steve: for the standard?

ralph: ideally yes - the normative materials the tf needs to know

ACTION: steve email 13250 to the WG

Task Force Updates

PORT

Alistair:quick start in response to action from f2f
... there is an example
... using skos core in rdf/xml and n3 (following guus suggestion)
... recommends assigning uri's for concepts
... should have metadata about the thesaurus itself
... links to main docs
... and thats it
... If this is the right sort of document, should we do a WD?
... I would like to publish the quick guide document and skos core vocab
... as soon as is possible
... haven't produced a wd before
... need guidance and advice

<aliman> http://www.w3.org/2004/02/skos/core/spec/

david: you'll get help

ralph: you have danbri

steve: subject property indicator - is there an example

alistair: this property has only just been added
... its in the spec document
... you can launch and example from there
... of using subject indicators
... I've left it out of quick guide doc
... because I've tried to scope to standard thesaurus terms
... and link to the longer document
... which includes discussion of different ways of identifying thinks
... the idea is to have an inverse functional property that refers to a psi docuement

steve: I'd like to ask about ...

alistair: that exert is a relative uri - the full uri is ....
... the full uri is the one they used in their publication of their thesaurus

steve: I'll take other questions to the list, specifically on use of xml:base

alistair: one question ...
... in the examples I chose to use an xml base to avoid repeating uri's
... with the possibility that people might miss the xml base
... comments on this style please
... bearing in mind that audience don't know rdf at all
... please send me comments
... I'd appreciate positive comments too

OEP

evan: I don't think anything has happened since the f2f. there is an agenda item proposal for a SE tf

evan: my question concerns the SE engineering tf formation

<Guus> Natasha and Alan gave a great tutorial at ISWC

Wordnet

Brian: I have some progress to report
... thanks to Andreas for his actions
... we've made some progress on the technical aspects of the ontology
... I have recruited some help from a student working at HPlabs
... he's fixed some bugs and is using Protege to make some OWL statements about the Wordnet concepts
... a number of issues have arisen
... I've been trying to have a document that an RDFschema-only processor can make use of
... and obviously an OWL processor would do more with this document
... would not model all the constraints in the Wordnet structure
... and OWL document would model more of the Worndet constraints
... thinking of something with an RDF Schema up front and Protege-generated OWL statements at the back
... I'd like feedback on this approach
... we have discovered that when you combine RDFS and OWL, Protege is not happy with the result
... I hope to post something to the list tomorrow

<aliman> protege OWL plugin is still pretty buggy in my experience; lots of things can throw it off.

XML Schema datatypes

jeff: I've discussed iwth jjc
... we have two new actions; one about duration

evan: has question re duration issue
... you said you would put something in about durations

jeff: jjc has sent email about adding a new section about duration

david: evan you can ask on list

ralph: has he sent it yet

jeff: yesterday

vocab mgmt

tom: we discussed in f2f which vocabs would be featured
... foaf and dc are in
... what about skos?
... alistair you suggest that skos illustrates some of the good practices

alistair: I'm happy for it to be used if other folks are happy iwth that

tom: lets put it in and review in draft
... re wordnet
... its not going to be ready
... part 2 has practices like use uri references
... and part 3 where things are less clear
... since wordnet is not a maintained vocab in the same sense as others
... didn't seem like a candidate for part 2
... but could look at it for part 3
... does anyone have a strong opinion
... shame aldo isn't here today
... he did volunteer to produce some info
... about practice in the context of wordnet
... but I'm assuming that at this point this will go into part 3
... I've been in touch with prism vocab maintainers
... they are a good candidate
... they exemplify principles of good practice
... they are looking to see if they have a w3c member
... otherwise I was proposing them I work with them to put in information about prism as appropriate with their help
... comments or objects?
... prism is a dc based vocab for print and magazine publishers
... there still is a need for a candidate vocab for a larger scale thesaurus or ontology
... that could illustrate some of the principles of good practice

TomB: there is one at FAO (fisheries?)
... its not clear we have one that is ready

ralph: I wondered specifically about oasis published subjects
... have you had a chance to consider that

tom: oasis published subjects is already in there
... do you mean as a thesaurus

ralph: as an example of vocabulary that will be maintained
... I was thinking of f2f discussion
... we don't want to point to people whom we are not confident will continue to follow best practice
... bringing oasis into a discussion aboout sw best practice might have other good effects

tom: we already have them in the introduction
... but that is not looking at them as a thesaurus.
... its already in there

alistair: if you used published subjects that would be an example of identifying terms indirectly
... and there is nothing in the draft about that
... we'd have to expand the document
... second thing is I've just posted a couple examples of large thesauri that have published in RDF
... they are not maintaining as an rdf vocab
... they are conversions from other forms of vocabs

ralph: I want to distinguish different aspects of why published subjects may be interesting
... they are using some specific techniques that are out of scope for vm
... but contrasts with wordnet which is large and has a maintenance activity
... which we are unlikely to be able to influence
... but published subjects may have a less well established maintenance process
... our practice should be independent of semantics of vocab

rdf in xhtml

David: I'll mention thtat there have been several messages on list
... from jjc and mark, also from ben
... big question is whether html wg addressed rdf/a

ralph: we checked the web
... best we could find was the irc logs
... acknowledged our encouragement
... but no specific discussion
... they are moving to last call

david: did they note jjc's feedback
... jjc's feedback was substantial and on point

ralph: the message sent to the tf mailing list didn't have a lot of detail
... Mark and Jeremey clarified the issues in a one-to-one meeting
... didn't say they'd resolved the issues
... I'm asking if there can be more detail
... I didn't get a warm cosy feeling that resolutions would appear in the last working draft
... mark has said there is not a lot of work to be done

david: should we take an action to follow up more directly

ralph: I will be asking steven permberton for more detailed records of their meeting
... as a practical matter, that wg is trying to go to last call this month

ACTION: ralph contact steve pemberton to clarify html wg's position on inclusion of rdf/a in their last call wd.

adtf

<DavidW> Tom Adams' notes on Tutorial Page

Phil's proposal for Software Engineering task force

ralph: lets postpone because of time

ACTION: david put software engineering task force on agenda for two weeks time

phil: please send feedback on draft terms of reference