DCMI Usage Board - Wednesday, 2008-05-28 2100 UTC telecon - report

This report:  http://stage.dublincore.org/usageboard/log/2008/2008-05-28.dcub-telecon-report.html
UB meetings:  http://stage.dublincore.org/usageboard/log/
UB mail list: http://www.jiscmail.ac.uk/lists/dc-usage.html
UB wiki:      http://dublincore.org/usageboardwiki/
Agenda:       http://stage.dublincore.org/usageboard/log/2008/2008-05-28.dcub-telecon-agenda.html
              http://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0805&L=dc-usage&P=1101

Attended: Stefanie, Joe, Pete, Akira, Andrew, Tom
Regrets: Julie, Diane

----------------------------------------------------------------------
Range of dcterms:title and new term proposal (dcterms:titleAsText)

Proposal for new term dcterms:titleAsText (Akira)
-- http://dublincore.org/usageboardwiki/TitleProposal
   http://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0805&L=dc-usage&P=1467

Options discussed:

1) dc:title remains without range.
   dcterms:title remains without range.
   No new parallel property.
   Pro: Same property for both literal and "sequence of words".
   Con: Providers must choose between literal value and
        non-literal value. Complex for consumers, OWL-DL problems.

2) dc:title remains without range.
   dcterms:title gets literal range.
   dcterms:titleAsSequenceOfWords gets range dcterms:SequenceOfWords (two new terms)
   Pro: Clear (one property per pattern, with dc:title ambiguous).
        No complexity-of-processing or OWL-DL problems.
   Con: Providers must choose.

3) dc:title remains without range.
   dcterms:title gets literal range.
   No new parallel property.
   Pro: Simple for consumers and providers, no OWL-DL problems
   Con: Transliteration case must be handled with dc:title or with 
        a newly coined, non-DCMI property.

4) dc:title remains without range.
   dcterms:title gets range dcterms:SequenceOfWords (one new term)
   Pro: Straightforward, no OWL-DL problems.
   Con: Makes use of dcterms:title with literal value invalid.

Resolved: to postpone consideration of the proposed new term and
to de-couple the creation of a new parallel property from the issue
of a literal range for dcterms:title.

ACTION 2008-05-28: Tom to propose literal range for
dcterms:title for finalization on the list.

Discussion:

-- The proposal would actually require the creation of two
   new terms (an additional proposal for the range class
   dcterms:SequenceOfWords).

-- The proposed approach would not cover the case of a single
   title with translations.  The more general requirement
   seems to be that of treating the title as a resource.
   But we have already seen that this requirement can come up
   not just for titles, but for just about any property with
   a literal range - at any rate, for those properties that
   are typically used with sequences of words.  This seems
   like a general issue requiring a general solution.

-- The explanatory overhead would be particularly high for
   this proposal, in part simply because it is a parallel
   property (which needs to be explained) and in part because
   the approach is more specific than simply treating title
   as resource (which would accommodate the case of a title
   with multiple translations or a title that needs to be
   annotated in other ways).

In the meantime, the requirement of multiple lexical forms
for transliterations can currently be handled using dc:title,
with its unspecified range.  It is true that this is a complex
range, but it is no worse in this respect than the cases of the
SKOS "note" properties or dcterms:description, which also 
currently have no assigned ranges.


-- 
Dr. Thomas Baker <[log in to unmask]>
Director, Specifications and Documentation
Dublin Core Metadata Initiative