W3C home > Mailing lists > Public > public-swbp-wg@w3.org > November 2005

Afternoon SWBPD OEP F2F session 2005-11-05 minutes

From: Elisa F. Kendall <ekendall@sandsoft.com>
Date: Sat, 05 Nov 2005 14:56:10 -0800
Message-ID: <436D388A.8060300@sandsoft.com>
To: public-swbp-wg@w3.org


OEP
--

Chris Welty - First note: N-ary relations, question is whether it should go to
   working draft or note status
Ralph - Thinks it is appropriate for working draft; WG needs to discuss whether or not 
   another version is anticipated in the lifetime of the working group
Jeremy - Give it note status
Chris - I'm happy with that. Publish it as a note, though I thought the process was to 
   go to working draft for some period of time first.
Ralph - If this were going to be a recommendation, that would be appropriate, but
   in this case we can just go directly to a note.
Guus - Do you want to call now for a vote?  The version we should be looking at ...
Chris - I put the URL in the IRC, or go to the OEP TF page - first note under editor's 
   draft, Version of August 11th 2005
Guus - Proposal to move this document to be a note. Straw poll needed? (no response)
   Made motion to move this document to note
DanBri - second
Evan - Second
Guus - No objections 
Jacco and Alistair abstain
Guus - No objections; so resolved
*** ACTION on Chris and Alan to work with Ralph to make it ready for publication
 by 1 December

Chris - Next note is Qualified Cardinality Restrictions, first drafted by Guus
   about a year and a half ago.  Alan took it over this summer and responded to his 
   own review, now an editor's draft.
   New draft will be on as soon as Chris pushes it through CVS, needs reviewers
Jeremy - had seen what he thought was a rough sketch
Alan - includes more detail / use cases now
Guus - hadnt been done sooner because there were process issues related to OWL working 
   group, thought it was important to raise the issue as a note even though the changes 
   didn't get into the OWL specification; They should be made known for the next revision.
Chris - *** ACTION: will put up tonight from his hotel room and needs reviewers ...
   To do items were put there as a place holder for this meeting, but *** ACTION: Chris
   will move those to the changes section
*** ACTION: Alistair will review (Has read it already, not expecting too much work) by 11/25.
Guus - that date makes it doable for discussion on our next telecon
Jeff - *** ACTION: will also review
Mike - Process comment - should be possible for someone else in the task force who had
   nothing to do with the note to review
Guus - not willing to change the process at this date, rationale is that someone
   outside the group can grasp the content if outside reviewers can do so.
*** ACTION: Jeff is still on the hook
*** ACTION: Jacco is also willing to review from a non expert perspective

Chris - Next note: OWL Time Note
Chris - We have an outstanding review due from a WG chair, Libby has done her review
    Comments were addressed
Libby - is satisfied with the result
Chris - There are actually two notes -- one OWL Time, the other Time Zone
Ralph - missed Libby's review (maybe a month ago)
    question - since the IG published a calendar note, the community should have done some 
    homework on the relationships between the two; Feng is working on this
... Also a comment made by someone regarding legal HTML, time work going on in that group
... Internationalization WG published a note a couple of weeks ago on time zones that
    should be referenced -- Libby brought that up in her review
Jeremy - time note uses the duration datatype, which is not a best practice, and
    should not be used; the XSLT work has solved the duration datatype issue
*** ACTION: Jeremy action to send comments to the list on the non-use of the duration datatype
Jeremy - technical question regarding default system political region - semantics can
    only appear in documentation, not in OWL
Chris - yes - there are axioms for this, and hopefully once there is a rules language
    the semantics can be implemented in that (note does not say that);
    the note will express the semantics that can be expressed in owl and remainder in text
    I don't know what will happen after the working group, so we should not commit to using 
    the rules language at this point ...
Evan - looking for a sentence that says they should be implemented once there is a 
    rule language sufficient to implement them
Mike - too dangerous
Chris - also didn't agree with it
Evan - *** ACTION: to send note to Feng on this, so that the comment is at least provided to them
David Booth - looked at the time zone document only, not clear to him why RDF was
    appropriate for this problem, a simple table look-up technology might be more appropriate
    there are two phases to the problem - identifying the time zone, which depends on the region, 
    and then looking up the algorithm to apply to compute the appropriate time; not clear that 
    RDF is the appropriate solution; if there is a good reason for doing so, it isn't clear in 
    the note
Chris - if you want to work the notion of a time zone into a larger solution that has a
    temporal component, and you want to be able to reason about time for some purpose, this is
    what you would use; not written because the world needed an RDF time zone (standalone)
    solution but because it was needed for reasoning about time
David - would develop an OWL solution that only used UTC, and pushed this problem off
Chris - that's what the other note does
Jeremy - best practice on the web is not to assume time; also best practice to always
    specify the time zone, should be marking things up with UTC offsets explicitly
Chris - yes, but there are cases where it may be needed, and this is simply a proposal
    for how you should use it if it is relevant to a particular application;
    Alot of the initial work on this came from annotating text about information where the
    time zone is often implicit
Deb - so this is the only emerging standard I'm seeing for this kind of thing
Chris - Libby did point us to another one, this was funded by the DAML program, and it's
    quite old, at the time the group was led by Jerry Hobbs, and an extensive review of all
    time proposals available at the time was done, and authors were involved at that point
    What we're doing here is taking that work and turning it into OWL
Guus - hasn't had an opportunity to review it, but why don't the datatypes for datetime
    in xml schema datatypes address this
Chris - isn't familiar with the DateTime datatype
Jeremy - time zone is either implicit or specified, but some documents don't specify it,
    and this provides a means for that
Chris - the time zone ontology provides the notion of a geographical region, which has some
    notion of mapping from region to UTC time associated with it, and it has all of the instances
    all of the time zones are specified, allows you to specify more than that
Libby - some calendaring applications require this; the W3C note talks about why time zones, 
    independently of the offsets, may be needed and are useful
Jeremy - what other examples of ontologies are there?
Chris/Guus - SKOS, units of measures
Jeremy - I'm more comfortable with patterns coming out of this group rather than this note
Alistair - As a user of this, I'm just looking for something that works
Chris - feedback on earlier notes that had lots of options was "tell us which one to use"
    The work on which this note is based has already undergone extensive review, so I'm very
    comfortable with it.
DanBri - make list of the namespaces the task force is using
Chris - the task force should mnaintain a registry
Guus - we as a working group should propose a strategy for that - dan might be in a 
    position to make a proposal here; this issue has come up over and over again
    This is the first one we are going to publish as an ontology - should it be a W3C namespace,
    for the ontology itself; what should be the base URI
DanBri - the W3C webmaster's will figure that out
Guus - should it be under best practices, or somewhere else
DanBri - you can find recommendations and notes in the TR space; should there be a note
    shell around this that would be in the TR space
Guus - there has been some discussion on this and whether there should be a separate
     ontology space
Ralph - as to the choice of the namespace URI, the editors get to propose something
Chris - there should be a standard pattern for ontologies
Ralph - this is something that has been discussed - you want a way for people to browse
     ontologies (consistently), we haven't been presented a compelling reason to provide such
     a naming convention, the default would be something like /2005/11/time
     if this working group wants to state a position on how the W3C maintains a uri space
     that can be used for this ... this shouldn't hold us up from going forward
Chris - this is something we would like, but it isn't that important
Guus - we have exactly the same problem with units of measures
DanBri - take this to the coordination group ** action to Guus **
Chris - any more comments on the time notes
Ralph - appreciates the fact that this is based on prior work, well informed, there is
    evidence that there is a community that has paid attention to this, but unlike SKOS
    there isn't evidence that it is widely used by folks who have worked with OWL and RDF ...
    if there is evidence of that, or of how it has been used, we should say that
Chris - i believe it says that
Ralph - Feng has not been appointed to the working group, so 
    *** ACTION: Chris is taking action to get ISI to appoint him
    Also, documents should not say that they are working drafts if they are only editor's drafts
DanBri - there is a template now for editor's drafts
Guus - action for Dan to remind everybody on the list where the link is (bottom of home page)
    coordinators of OEP to remind editors regarding use of "working draft"

Chris - First editor's draft for Semantic Integration and Interoperability editors's draft
    Mike Uschold/Chris Menzel
Mike - looking for feedback from the WG for general reaction, requirements, scope issues
    looking for reviewers
    3 things for reader to take away -- aware of situations requiring interoperability, 
    understand the constructs form OWL and RDF that can be used to express mappings, and
    be able to create simple mappings for their own ontologies
    Review outline of note 
    Overview of use cases 
    Principles related to supporting interoperability
Deb - statements in owl were provided in the document, you might want to express more
    interrelationships but this will get you started, do have an examples of some things in KIF
    that you're not going to be able to express in OWL
Chris - the core thing is to express the relationship between the meaning of a term in one 
    ontology to the meaning of a term in another ontology 
Phil - possibly two notes in here using owl as a canonicalization for expressing ...
    there is a piece of work before this note - this is too specific; there are different
    contexts in which you want to glue/map different information constructs; owl provides a 
    mechanism for doing this in the web;
    Potential intro to say that when we talk about interoperability there are different
    architectural approaches - mapping could be one, canonicalization in and out of owl might 
    be another
Chris - Michael Gruninger and Mike Uschold have a paper about this
Jacco - reusing existing ontologies should be first over creating new ones
Mike - Back to the note, use case - my airline ontology vs. your airline ontology
    mapping examples - will refer to papers that cover the various examples, a catalog of mappings, and how they are used
    examples -- class mappings (show how OWL can support more complexity)
Raphael - only showed equivalence relations (document will include more)
Deb - Document is behind the presentation
Mike - issues and limitations, missing constructs needed that are not in OWL (string manipulation,
    computation, rules)
Guus - these kinds of semantic integration statements are very important, it's an extremely
     difficult area, worried that given the lifetime of the group we will be able to get consensus;
     would be in favor of something that is based on an extension of the lifetime of the working
     group
Phil - thinks there should be at least a separate task force if not a separate working group
    to address this
Raphael - saying what OWL can do/not do is still good information
Fabien - title needs to narrow the scope, potentially
Alan/Deb - "Towards ..."
Jeremy - It sounded like the editors have the goal of a narrow scope, thus it may be
    achieveable given the narrow scope
Alistair - supports it, but would like to see it factored out and scope tightened, would like
    to have, in the vein of the other OEP notes, practical, "I've got some RDF and OWL, how do I
    deal with semantic integration" -- not try to address the whole problem, 
Guus - are you saying to take a case study approach
Alistair - I have some knowledge problem that I don't use RDF/OWL for currently - take that 
    out for now; might add something about practical consequences for reasoners, how scalable, 
    data translation
DanBri - techniques instead of 
Phil - there are alot of people interested in this area, slightly concerned about a novice
    reading this type of note and not picking up on the depth/breadth of the problem
Mike - We'll be very clear about that in the issues and limitations
Jacco - complaint he has heard from the DB community, that the semantic web people are 
    reinventing all of the work we have done on schema integration
Deb - we should look at that
Evan - change to alignment of instance data for reasoners; different use case from data 
    integration/translation
*** ACTION: Reviewers - Raphael Georgios S, Fabien, Phil
Guus - plans for the future - for the task force, and this note
Chris - if you look at the TF web page, we had all kinds of plans initially for lots and
    lots of notes, there is a long section of suggested topics, suggested by people outside, 
    including an OWL version of Ontology 101, common pitfalls
Guus - ontology 101 should go under education and outreach
Chris - there is a lot of work to be done, some of which falls under education and
    outreach, which alot of people consider to be really basic, good examples in a namespace
    where people expect it to be is pretty important
    if the working group continues then OEP should be continued, if not then OEP should
    be continued some other way;
    on the note - Chris Menzel has joined the working group and has been very productive,
    working actively on this for the last couple of months, optimistic that this will get to
    a reasonable form by the end of the WG
Guus - it would be nice if there is a place to go if it isn't; to have a priority list of
    things that should be done
Chris - other notes planned include: 
    semantic integration
    units of measure
    space, spatial information
    fluents (may go - Chris has a draft)
Guus - so, at least 3 and possibly more notes planned, where there is outside interest
Chris - 13
Guus - a large number
Mike - there is another note that should be written, a potpourri of all kinds of simple things ...
Chris - pitfalls and tricks of the trade
DanBri - lat lon example, maybe an incubator for making a slightly broader version of that ...
    simple RDF vocabulary for doing something like that - perhaps this group [OEP] is the right 
    home
Chris - there is a huge community, with a propensity for not reusing old work -- we shouldn't
    consider new work in that area without considering what has been done; this would be a good
    place, goes along with what OEP has done
DanBri - give us something simple vs. take a huge spec and implement it in RDF/OWL, huge
    scope difference to consider
Chris - the work is large enough to be its own working group - building a standardized
    ontology for geographic information; Tony Cohen agreed to be involved but he's at the
    University of Leeds, they are very small, so they are unlikely to join the W3C; he can't
    do the work if he has to join the W3C
Alan - can put people together who are interested in this
Mike - there will be folks at thte owl workshop who might be interested
DanBri - *** ACTION: will take an action to investigate this further, follow up on spatial and geo
Chris - thinks it is big enough to be a working group, but most of the people are outside
    of the W3C at this point

Guus - each task force leader to put plans on the white board, what is a sensible way forward
    scenarios for rechartering, extending, starting new groups - with a clear set of proposals 
    for doing so 
Ralph - kudos to OEP for documenting ideas on their home page
Jeremy - proposed a 1/2 hour extension, Chris seconded
    future is too important to cut the meeting short, ....agreed
Received on Saturday, 5 November 2005 22:56:21 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:17:19 GMT