[Esip-preserve] Harmonizing data citation guidelines

[Mark A. Parsons]

Hi all,

I started reviewing the few data citation guidelines I was aware of  
plus the several Bob Cook sent around a while back. Below is a quick  
and dirty assessment.

What other guidance is out there? How do they compare to these?

IPY: http://ipydis.org/data/citations.html
      Detailed guidelines on how to construct a citation in the manner  
of the Chicago Manual of Style. The idea is to provide all the  
elements that any journal might require and to link to the data  
unambiguously. It is very much geared toward collections rather than  
specific files. It doesn't handle versions particularly well.

ORNL: http://daac.ornl.gov/citation_policy.html
      Doesn't include all the detail of IPY but follows the same basic  
principles. More emphasis on the DOI. Includes a link to an nice essay  
by Bob providing justification.

NSIDC
Has a similar approach to IPY, but isn't as rigorous about dates used.  
We have not yet adopted DOIs. We do not have publicly published  
guidelines, but all data sets provide an example and request to cite.

Pangaea: http://www.pangaea.de/about/
      Very much in line with IPY. Strong emphasis on DOI, which leads  
them to wait until a data set has reached some level of completeness  
before providing a citation. They tend to provide DOIs for fairly  
small data sets, which other data centers might compile into a larger  
collection. I believethis is an attempt to mediate the versioning  
problem.

GBIF: http://www.gbif.org/participation/data-publishers/gbif-sharing-agreement/how-to-cite-gbif-data/white-paper-citation-of-gbif-data/
      Under construction? I thought they used to have something.

NISO:  http://www.niso.org/topics/tl/NISOTLDataReportDraft.pdf
      A good exploration of the issues, but not specific guidelines  
(calls for guidelines)

DataVerse:  http://thedata.org/citation
      These folks are gears toward social science, but they have  
something going on.  They have something called a universal numerical  
fingerprint (UNF) that automatically changes whenever any part of the  
data changes. It "is designed to persist even if URLs--or the web  
itself--are replaced with something else.". They also have an  
automated system to create a citation for a data set, i.e the UNF

TDWIG:  http://www.tdwg.org/standards/150/
      Not a citation approach, per se, but good, detailed guidelines  
on the use of unique identifiers--the LSID in this case.


-m.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.lists.esipfed.org/pipermail/esip-preserve/attachments/20091110/0af563d8/attachment.htm>