[Metadata] CALL FOR CONUNDRUMS!

[Norman Paskin <N.Paskin@DOI.org>]

Call for Conundrums!

As many people on this list know, I have spent a considerable time over
the past weeks attempting to putting into place a robust metadata
framework for the DOI.  This has meant a quiet time on the lists in the
absence of real material.  We now need to test the metadata proposal we
have made and we need your help.  This is a great opportunity for you to
throw anything you?ve got at it!
 
We have recently published a version 3 of the DOI metadata paper,
dealing with DOI metadata structure; and a paper on DOI metadata
principles dealing with issues such as DOI registration and flow.  I will
soon be publishing a brief proposed practical implementation guide for
one example genre (probably in the area of scientific journals: a simple
set of rules to follow for DOI registration with metadata); but you need not
wait for that to give us CONUNDRUMS now..  

This call for conundrums relates solely to DOI metadata structure issues
(not, at this stage,  registration preocedures).  We want REAL
EXAMPLES of difficult publishing problems the DOI will have to deal with. 
How do I give a DOI to this..should there be one DOI or two...how do I
decide what to do with this... As our first challenge might be reference
linking between articles, I am interesting particularly BUT NOT ONLY in
such examples.  Examples could be:
- A paper in print form, and an electronic equivalent that also carries
addditional electronic-only material
- A paper that?s published simultaneously by two journals, by agreement
- A preprint server and a journal which produces equivalent papers ?for
the record?
- A paper that is a conference report containing several abstracts of
other papers
- An encyclopaedia on line thats updated continually
- A journal article in several different formats published at different times
- A journal article citing an archive source, an electronic structure
database, etc
- material in some form of supplement archive
- and many more we no doubt have not thought of.

In order to make this as real as possible, we want more than hypothetical
examples (though hypotheticals will be dealt with too).  We want real
examples and bibliographic details and background so that we can show
how to make real metadata records that will be persistent, following just
some very basic rules.

Please send the problems to the discuss-doi list, or (especially if they
have a lot of detail) directly to me.

Reference linking is a challenge here, but also, please think one step
beyond, as we think we have a basic infrastrucutre model for other
applications. If the consequences look difficult, thats because it is a
difficult subject to explain in the abstract, but there is no better test than
real examples.  (Recently Adam Phillipidis  posted a query about preprints,
which was one such useful conundrum). 

As I have noted elsewhere, for those still having concerns about DOIs
use of/  alignment with INDECS, may I point out that the INDECS
framework at www.indecs.org is a practical guide to solving such
conundrums.  Thats all our alignment means.  Please give this your best
shot!

Norman 


------------------------------------------------------
Metadata maillist  -  Metadata@doi.org
http://www.doi.org/mailman/listinfo/metadata