Title:         Meeting notes, Usage Board face-to-face, Singapore 25-26 August 2007
Identifier:    http://dublincore.org/usage/minutes/2007/2007-08-26.dcub-meeting-notes.html
Source:        e:/work/dcublog/2007/2007-08-26.dcub-meeting-notes.txt

Present:    Joe Tennis, Tom Baker (chair), Stuart Sutton, Andrew Wilson, Akira Miyazawa, 
            Diane Hillmann (Day 2)
Guests:     Raju Buddharajua, Pete Johnston, Traugott Koch (Day 1), Fredrik Enoksson (Day 2), 
            Mikael Nilsson (Day 2)
Ex officio: Makx Dekkers

Agenda
-- http://dublincore.org/usage/meetings/2007/08/singapore/

Meeting packet
-- http://dublincore.org/usage/meetings/2007/08/singapore/2007-08-25.ub-meeting-packet.pdf

Raw meeting notes
-- http://dublincore.org/usage/minutes/2007/2007-08-26.dcub-meeting-notes-raw.html


======================================================================
1. Changes to Terms in the DCTERMS Namespace (Diane) [lead by Tom in Diane's absence]

-- From the agenda:
   http://dublincore.org/usage/meetings/2007/08/singapore/.html/dcterms-changes.html - topic page
   http://dublincore.org/usage/public-comment/2007/07/dcterms-changes/

-- Errors noticed
   1.  Types of Encoding Schemes --
   2.  Block on third page of the document -- formatting screwed up, missing 4646
   3.  4646 -- datatype vs. VES needs to be resolved
   4.  in paragraph "New Terms Related to the Abstract Model" - change DCAM "is member of" to a "member of"
   5.  section 2 -- change "free-standing names" to "free-standing labels"

    ACTION 2007-08-25: Tom to revise Introduction to "Revisions to DCMI Metadata Terms",
    correcting errors noted.

-- Section 3 Properties

    dcterms:creator -- APPROVED
    dcterms:source -- APPROVED
    dcterms:coverage -- APPROVED
    dcterms:language -- APPROVED
    dcterms:subject -- APPROVED [noted, the list of examples is incomplete]
    dcterms:title -- APPROVED
    dcterms:audience -- APPROVED
    dcterms:alternative -- APPROVED
    dcterms:tableOfContents -- APPROVED
    dcterms:abstract -- APPROVED
    dcterms:created -- APPROVED
    dcterms:valid -- APPROVED
    dcterms:available -- APPROVED
    dcterms:issued -- APPROVED
    dcterms:modified -- APPROVED
    dcterms:extent -- APPROVED
    dcterms:medium -- APPROVED
    dcterms:isVersionOf -- APPROVED
    dcterms:hasVersion -- APPROVED
    dcterms:isReplacedBy -- APPROVED
    dcterms:replaces -- APPROVED
    dcterms:isRequiredBy -- APPROVED
    dcterms:requires -- APPROVED
    dcterms:isPartOf -- APPROVED
    dcterms:hasPart -- APPROVED
    dcterms:isReferencedBy -- APPROVED
    dcterms:references -- APROVED
    dcterms:isFormatOf -- APPROVED
    dcterms:hasFormat -- APPROVED
    dcterms:conformsTo -- add the word "described" therefore it should
       read: "An established standard to which the described resource
       conforms" APPROVED  

    dcterms:spatial -- APPROVED
    dcterms:temporal -- APPROVED
    dcterms:mediator -- APPROVED
    dcterms:dateAccepted -- APPROVED
    dcterms:dateCopyrighted -- APPROVED
    dcterms:dateSubmitted -- APPROVED
    dcterms:educationLevel -- add 'described' and change 'whom' to
       'which' thus, "...for which the described resource is intended"
       -- APPROVED
    dcterms:accessRights -- APROVED
    dcterms:bibliographicCitation -- APPROVED
    dcterms:license -- Tom to explicitly say "New Comment [comment
       deleted]" -- APPROVED
    dcterms:rightsHolder -- APPROVED
    dcterms:provenance -- APPROVED
    dcterms:instructionalMethod -- APPROVED
    dcterms:accrualMethod -- APPROVED
    dcterms:accrualPeriodicity -- APPROVED
    dcterms:accrualPolicy -- APPROVED

-- Section 3 Vocabulary Encoding Schemes and Syntax Encoding Schemes

    DISCUSSION: two issues were raised [TK]
       1. are all vocabularies defining? -- many are just lists, labels, or
          the like, with no formal definitions
       2. what is the relationship between concepts and terms in vocabulary
          control?

    AGREED: we can use 'specified' instead of 'defined'
    Options for fixing #2 above:
       * set of concepts specified by...
       * vocabulary specified by...

    AGREED: change dcterms:LCSH to the set of labeled concepts
    specified by Library Congress Subject Headings

    RESOLVED 2007-08-25: In all Vocabulary Encoding Schemes
    change 'defined' to 'specified'

    dcterms:LCSH - change dcterms:LCSH to 'the set of labeled concepts
       specified by Library of Congress Subject Headings' -- APPROVED with change

       [Corrected in 2007-11-23 telecon to read:  
       "specified by the Library of Congress Subject Headings"]

    dcterms:MESH change dcterms:MESH to 'the set of labeled concepts
       specified by the Medical Subject Headings' - APPROVED with the change noted above

       [Corrected in 2007-11-23 telecon to read:  
       "specified by the Medical Subject Headings"]

    dcterms:DDC -- 'the set of conceptual resources specified...' -- APPROVED

       [Corrected in 2007-11-23 telecon to read:  
       "specified by the Dewey Decimal Classification"]

    dcterms: LCC - 'the set of conceptual resources specified...' -- APPROVED

       [Corrected in 2007-11-23 telecon to read:  
       "specified by the Library of Congress Classification"]

    dcterms:UDC - 'the set of conceptual resources specified...' -- APPROVED

       [Corrected in 2007-11-23 telecon to read:  
       "specified by the Universal Decimal Classification"] 

    dcterms:DCMIType -- 'The set of classes specified by the DCMI Type
       Vocabulary, used to categorize the nature or genre of the resource.'
    dcterms:IMT -- 'the set of media types specified by the Internet
       Assigned Numbers Authority' - APPROVED
    dcterms:ISO639-2 -- 'The three-letter alphabetic codes listed in
       ISO639-2 for the representation of.....' APPROVED
    dcterms:RFC1766 -- APPROVED
    dcterms:URI -- change 'defined' to 'specified', write out
       Internet Engineering Task Force, also: could take clarification
       somewhere, but should remain consistent between DCMI usage and XML
       Schema -- APPROVED as changed

    dcterms:Point -- APPROVED
    dcterms:ISO3166 -- APPROVED
    dcterms:Box -- APPROVED
    dcterms:TGN -- APPROVED
    dcterms:Period -- APPROVED
    dcterms:W3CDTF -- APPROVED
    dcterms:RFC3066 -- 'The set of tags constructed according to RFC 3066
       for the identification of languages.' -- APPROVED with changed
       punctuation
    dcterms:NLM -- 'The set of conceptual resources specified by the
       National Library of Medicine Classification' -- APPROVED

    dcterms:RFC4646 - 'The set of tags constructed according to RFC 4646
       for the identification of languages.' -- APPROVED with change in punctuation

-- Section 3 DCAM terms
    dcam:memberOf -- APROVED
    dcterms:VocabularyEncodingScheme -- APPROVED

    ACTION 2007-08-25: Tom to change introduction discussing 4646 to Syntax Encoding
       Scheme -- it currently reads Vocabulary Encoding Scheme

-- Issues arising

    ACTION 2007-08-25: (Someone?) to post a known list
    of issues to the UB list (1) for subject -- if read
    as inclusive, the list of examples is incomplete, (2)
    classification codes is misleading because it misses
    classification captions (use classification expressions
    instead), (3) 'spatial' and 'temporal' are "correct"
    but 'coverage' is too broad -- it allows for topic and
    this is in conflict with 'subject' [TK] -- but this is
    an Application Profile issue [TB].

    DISCUSSION 2007-08-25: because of the problem between
    'coverage' on the one hand and 'spatial' and 'temporal'
    on the other. We have two problems:

       1.  overlap between subject and coverage
       2.  subproperty relations between spatial/temporal and coverage

    we have four courses of action:

       1. delete assertion spatial/temporal subproperty of coverage [problem
          for legacy and namespace policy]

       2. change definition of spatial/temporal to fit coverage [problem for
          namespace policy]

       3. do nothing and accept that spatial/temporal are not really
          subproperties

       4. introduce a new coverage or recommend spatial/temporal instead

    NOTED: 2004 definition of 'coverage'
    was significantly different from 2007 -- and we may
    have introduced a problem in the change in 2007 i.e.,
    constraining 'coverage' by using the word 'topic'

    NOTED: DCMI might want to build an AP that
    ameliorates the legacy problems of 1.1 and TERMS --
    'coverage', 'identifier', 'source', etc. -- this needs
    to be linked to discussion and decision of the AP of
    Simple DC.

    AGREED: 'spatial' and 'temporal' subproperty
    assertions are inconsistent with the definition of
    'coverage'.

    ACTION 2007-08-25: Tom to put discussion of dcterms:coverage
    issues onto the telecon agendas.

======================================================================
2. Structure of DCMI Metadata Terms document (Tom)
   -- http://dublincore.org/usageboardwiki/DcmiTermsOutline?action=print - archived as
      http://dublincore.org/usage/minutes/2007/2007-08-26.dcub-meeting-topic02.html
   Changes to term declarations in RDF (Tom)
   -- http://dublincore.org/usage/meetings/2007/08/singapore/.html/rdf-declaration-changes.html

Question: How do we want to present DCMES, TERMS, DCAM terms?
-- Option #1 -- just list these in these sections
    1.1
    TERMS
    VES
    SES
    DCMITYPE
    DCAM

-- Option #2 -- put TERMS first, to emphasize our preference
    DCTERMS first
    Then 15 1.1 ns -
    VES
    SES
    DCMITYPE
    DCAM

AGREED: follow proposal #2 above

ACTION 2007-08-25: Tom to add clarification to Introduction
of http://dublincore.org/documents/dcmi-terms/ -- for example
rationale for replicating 1.1 terms in the TERMS namespace,
recommendations on what people should be using by preference,
and any other -- will circulate to AB for advice on anything
else that should be in that introduction

   http://dublincore.org/usageboardwiki/DcmiTermsOutline?action=print

ACTION 2007-08-25: Tom to prepare TERMS document structured
as follows:
    DCTERMS first
    Then 15 1.1 ns -
    VES
    SES
    DCMITYPE
    DCAM

ACTION 2007-08-25: Tom to add a note to the terms changes
document explaining change to the type of the terms in the
Type Vocabulary.

======================================================================
3. Review of application profiles (Joe, Stuart)

   http://dublincore.org/usage/meetings/2007/08/singapore/.html/profile-review.html - topic page
   http://dublincore.org/usageboardwiki/ProfileReviewCriteria?action=print    - old criteria
   http://dublincore.org/usageboardwiki/ProfileReviewCriteriaNew?action=print - new (Aug 2007) criteria
   http://dublincore.org/usage/meetings/2007/08/singapore/pdf-term-decision.pdf
   -- dublincore.org/architecturewiki/TermDecisionTree?action=print
   http://dublincore.org/usageboardwiki/SESandVES?action=print
   http://dublincore.org/usage/meetings/2007/08/singapore/.html/CollectionsProfileReview.html

-- Background documents on process
   http://dublincore.org/documents/approval/
   http://dublincore.org/usage/documents/process/
   http://dublincore.org/usage/meetings/2007/03/barcelona/Process_Doc_Revisions.txt

Usefulness -- requirements for the metadata are x,y, or z, and it is
documented as such

There are separate steps for deciding which APs to consider.

Criteria should be as clear as possible, and if possible, elicit a yes
or no answer to the question posed. Verbose explanations belong in a
separate document.

Usefulness should only be used in functional requirements relationship
to domain model.

ACTION 2007-08-25: Tom to add to the criteria in the wiki
http://dublincore.org/usageboardwiki/ProfileReviewCriteriaNew
a section called "Organizational Context": evaluation questions
to be asked about who has stewardship of a proposed AP.

======================================================================
4. FOAF (Andrew)
   http://dublincore.org/usage/meetings/2007/08/singapore/.html/foaf.html - topic page
   http://dublincore.org/groups/agents/agentFRdraft2-2.html
   http://dublincore.org/agentswiki/FoafReview?action=print
   http://dublincore.org/usage/meetings/2007/08/singapore/Eprints-excerpt.pdf
   http://rdfweb.org/mt/foaflog/archives/2003/07/10/12.05.33/

Does FOAF as a vocab meet the needs of the DCMI community? Does
DC have a role to play in providing an institutional context
for FOAF.

Dan Brickley proposed using FOAF for Agents Task Group.  Dan did
functional requirements of FOAF -- on Agent Task Group wiki.

Independent namespace reuse study
    Name
    Organization
    Website
    Email

FOAF is a private namespace -- so there are issues there

Two issues with using FOAF:
    1. desire to use FOAF in DCAM-based APs like Eprints
    2. FOAF could expire as a reliable namespace

Dan owns the namespace, but is not interested in submitting it to a
UB-like process.

Best option would be for the Agents TG/Community to decide
and forward an AP to the UB.  MN: need to decouple the FOAF
vocab from the AP issue.

ACTION 2007-08-26: Andrew, Tom, and Dan B, in context of Agents
TG, to finish the assessment of FOAF against the functional
requirements. Include context describing kinds of places
where FOAF would be useful and where it wouldn't be useful.
Following this assessment, the TG to propose/recommend a
course of action to DCMI Directorate.

======================================================================
5. Domains and ranges (Akira)
   http://dublincore.org/usage/meetings/2007/08/singapore/.html/domains-ranges.html - topic page
   http://dublincore.org/documents/2007/07/02/domain-range/

----------------------------------------------------------------------
Issues arising

ISSUE: Should we adopt 'non-literal resource' as a new range value?

Akira: All values will become 'literals'. A date written in the W3CDTF
is a literal but the date itself is *not* a literal. We need more
analysis of what ranges are and what values are appropriate.

Often a range value is just a rewording of a property. Pete says not a
problem.  Issue is whether the relationship between properties and
sub-properties is changed, not how the range restricts values. Choices
about property ranges made for APs affects eventual metadata
interoperability.

Akira: problem with relationship between ranges and properties.
Properties have meanings which are very close to the values of ranges.

NOTE (Tom): 'has domain of rdfs:resource' is an implicit assignment and
there is no need to declare it explicitly.

----------------------------------------------------------------------
Terms approved

dcterms:contributor -- APPROVED

    dcterms:Agent -- AGREED to change comment to read: "Examples of Agent
    include person, organization, and software agent." APPROVED

    dcterms:AgentClass -- change comment to read: "Examples of Agent Class
    include groups seen as classes, such as students, women, charities,
    lecturers." APPROVED.

dcterms:coverage -- APPROVED

    dcterms:LocationPeriodOrJurisdiction -- AGREED to change 'period' in
    definition to 'period of time'. APPROVED.

    dcterms:Location -- AGREED to change definition to: "A spatial region or
    named place." APPROVED

    dcterms:PeriodOfTime -- AGREED to change definition to: "An interval of
    time that is named or is specified by its start and end dates." APPROVED.

    dcterms:Jurisdiction -- APPROVED

dcterms:creator -- APPROVED
dcterms:date -- APPROVED. 'Yes': 5; 'No': 1.
dcterms:description -- APPROVED
dcterms:format -- APPROVED

    dcterms:MediaTypeOrExtent -- APPROVED
    dcterms:MediaType -- APPROVED
    dcterms:PhysicalMedium -- AGREED to change comment to: "Examples include
    paper, canvas, or DVD" and to change definition to: "A physical material
    or carrier". APPROVED
    dcterms:PhysicalResource -- APPROVED
    dcterms:FileFormat -- APPROVED
    dcterms:DigitalResource -- REMOVE because not used in any Domain or
    Range assignment. It was only proposed for reasons of symmetry with
    PhysicalResource.
    dcterms:SizeOrDuration -- AGREED The definition is to be redone.

ACTION 2007-08-26: Pete to propose new definition for 
the proposed class SizeOrDuration.

    dcterms:Duration -- REMOVE because not used in any Domain or Range
    assignment. It may make sesne to create such a subclass of
    SizeOrDuration at a future date but there may be better ways to subclass
    SizeOrDuration.

    dcterms:Size -- REMOVE because not used in any Domain or Range
    assignment. It may make sense to create such a subclass of
    SizeOrDuration at a future date but there may be better ways to subclass
    SizeOrDuration.

dcterms:identifier -- APPROVED
dcterms:language -- APPROVED

    dcterms:LinguisticSystem -- APPROVED

dcterms:publisher -- APPROVED
dcterms:relation -- APPROVED
dcterms:rights -- APPROVED

    dcterms:RightsStatement -- AGREED to change 'Resource' to 'resource' in
    definition. APPROVED

    dcterms:LicenseDocument --  AGREED to change 'Resource' to 'resource' in
    definition. APPROVED

dcterms:source -- APPROVED
dcterms:subject -- APPROVED
dcterms:title -- NOT APPROVED. Vote -- Yes: 3; No: 1; Abstain: 2. 
-- As a consequence the range for 'title' remains unspecified.

dcterms:type --  APPROVED

dcterms:abstract -- APPROVED
dcterms:accessRights -- APPROVED
dcterms:accrualMethod -- APPROVED

    dcterms:MethodOfAccrual -- APPROVED

dcterms:accrualPeriodicity -- APPROVED

    dcterms:Frequency -- APPROVED

dcterms:accrualPolicy -- APPROVED

    dcterms:Policy -- AGREED to change definition to: "A plan or course of
    action by an authority, intended to influence and determine decisions,
    actions and other matters." APPROVED

dcterms:alternative -- NOT APPROVED. See note above for dcterms:title
dcterms:audience -- APPROVED
dcterms:available -- APPROVED
dcterms:bibliographicCitation -- APPROVED

    dctermsBibliographicResource -- APPROVED

dcterms:conformsTo -- APPROVED

    dcterms:Standard -- APPROVED

dcterms:created -- APPROVED
dcterms:dateAccepted -- APPROVED
dcterms:dateCopyrighted -- APPROVED
dcterms:dateSubmitted -- APPROVED
dcterms:educationlevel -- APPROVED
dcterms:extent -- APPROVED
dcterms:hasFormat -- APPROVED
dcterms:hasPart -- APPROVED
dcterms:hasVersion -- APPROVED
dcterms:instructionalMethod -- APPROVED

    dcterms:MethodOfInstruction -- APPROVED

dcterms:isFormatOf -- APPROVED
dcterms:isPartOf -- APPROVED
dcterms:isReferencedBy -- APPROVED
dcterms:isReplacedBy -- APPROVED
dcterms:isRequiredBy -- APPROVED
dcterms:issued -- APPROVED
dcterms:isVersionOf -- APPROVED
dcterms:license -- APPROVED
dcterms:mediator -- APPROVED
dcterms:medium -- APPROVED
dcterms:modified -- APPROVED
dcterms:provenance -- APPROVED

    dcterms:ProvenanceStatement -- APPROVED

dcterms:references -- APPROVED
dcterms:replaces -- APPROVED
dcterms:requires -- APPROVED
dcterms:rightsHolder -- APPROVED
dcterms:spatial -- APPROVED
dcterms:tableOfContents -- APPROVED
dcterms:temporal -- APPROVED
dcterms:valid -- APPROVED

----------------------------------------------------------------------
Editorial actions and follow-up issues

ACTION 2007-08-16: Tom to edit "Domains and ranges":  in
cases where the domain is not specified add '[unspecified]'.

ISSUE: Whether we leave the range unspecified, or whether
we specify a range of 'non-literal'. This would require the
definition of a new class of 'non-literal'.

ACTION 2007-08-26: Tom to add a note to the terms documentation
for each term that has an approved range of rdfs:resource,
except for 'description' and its sub-properties 'abstract' and
'tableOfContents', to the effect that the intention is that
these terms are to be used with non-literal values, with a
pointer to the AM. The Usage Board is currently investigating
a model for expressing this intention formally.

ACTION 2007-08-26: Tom to add a note to the documentation for
'title' saying the intention is that while the Usage Board
recognises that title is used in practice primarily with
literal values we also recognise that important uses of
'title' with non-literal values and so the Usage Board is
investigating the options for expressing this.

ACTION 2007-08-26: Pete and Mikael to investigate the
usefulness of defining a class of 'non-literal' and assess what
the consequences of this would be for machine processability.
(Usage Board should track issues involved in the notion of
a range of 'non-literal'.)

NOTE: For dcterms:title, the range of "literal" was NOT approved.
For "parallel properties", DCMI could provide pointers to a
"parallel property" in another namespace from a DCMI vocabulary declaration.
Options:  1. Leave range unspecified
          2. Assign range of 'non-literal'
          3. Assign range of 'literal' [the proposal]

NOTE: Need for clarification of relationship between rdf:type and dcterms:type.

======================================================================
7. Description Set Profile model and wiki syntax (Mikael as guest)
   http://dublincore.org/usage/meetings/2007/08/singapore/.html/description-set-profile.html - topic page
   http://dublincore.org/usage/meetings/2007/08/singapore/pdf-dcap-model.pdf
   http://dublincore.org/architecturewiki/DescriptionSetProfile?action=print
   http://dublincore.org/usage/meetings/2007/08/singapore/pdf-eprints.pdf
   -- knowware.nada.kth.se/DCWiki/EprintsApplicationProfile?action=print
   http://kmr.nada.kth.se/wiki/Main/MoinMoinWikiFormatSuggestion1?action=print

======================================================================
8. Application Profile of Simple Dublin Core (Tom, Mikael)
   http://knowware.nada.kth.se/DCWiki/SimpleDublinCore?action=print

    Question being addressed is "What do we want to say simple DC is?". This
    work needs to be based on what functions we are trying to provide for.

    1. To provide an AP that tells implementors how to use the traditional
    '15' in a way that conforms to modern understanding of metadata
    descriptions based on the DCAM.

    2. Scope and purpose based on the mature understanding of the UB,
    developed over the last 6 years.

    To be explicit about a number of things, to comment, to point out flaws
    and do some house cleaning.

    Something that looks very much like the DC15; go into spatial and
    temporal; provide a pedagological clarification of the 'core'.

    The new AP is not intended to solve the general problem of 'resource
    description', which is best addressed by communities in Application
    Profiles.

    How to deal with source and identifier without going beyond the 15?
    NB We are not starting from scratch.

    Purpose (MN): to serve as a path for implementors of the simple DC to
    move towards a more modern implementation of DC metadata terms.

    Need to have a relatively conservative interpretation of the OAI-PMH
    version of DC.

    Need for a second profile - but what is its function and scope?

    Describe the legacy and the path to the new - in two profiles but
    presented as one package.

    Straightforward to create a AP for traditional 'simple DC'. Then...?

    A second AP that provides a modern way of implementing what people think
    of as 'simple DC' (but called something else).

    To reverse engineer traditional use of 'simple' DC.

    Reconstruct a new version of what was called 'simple DC' that implements
    modern understandings and ameliorates existing flaws in the DC15.

    APs or description set profiles?

    ACTION 2007-08-26: Tom and Mikael to create a draft AP
    using the 1.1 namespace and which models everything as
    literals. Document the legacy functional requirements and
    the organizational context for this AP.

    AGREED: Defer, for the time being, work on a second AP until we have a
    better understanding of the functional requirements, scope and purpose
    of this second AP.

2009-01-29: Changed usageboard/log URIs to usage/minutes URIs.