[ckan-dev] Citations in CKAN

[Joss Winn]

> 
> 
> Message: 3
> Date: Tue, 11 Sep 2012 22:07:44 +0100
> From: Ross Jones <ross.jones at okfn.org>
> Subject: [ckan-dev] Citations in CKAN
> To: CKAN Development Discussions <ckan-dev at lists.okfn.org>
> Message-ID: <E965737E-E33D-492F-9AE4-C20591BF461C at okfn.org>
> Content-Type: text/plain; charset="us-ascii"
> 
> Hello,
> 
> It's recently come up that someone has asked for citations on dataset pages in CKAN.  Whilst we're discussing the exact citation format I added a section to the dataset page on the datahub (see Cite this section on http://datahub.io/dataset/afterfibre).  It's likely that the link in the citation will also at some point contain the specific revision of the dataset (which is normally accessible via /dataset/name at revision_id) but beyond that nothing has been decided about how this should be implemented.  There was a previous conversation on the topic at http://lists.okfn.org/pipermail/ckan-discuss/2011-May/001236.html.
> 
> I think the best approach is to use the data hub as a place where we can refine the citations before we move it into CKAN itself, either as part of the new templates or as part of the component that also allows sharing via twitter etc.
> 
> If you have any suggestions, feedback or ideas relating to citations it would be great to hear them. I can't guarantee they'll get implemented immediately, but it should certainly be possible to point an interested volunteer in the right direction in the meantime.
> 
> Regards
> 
> Ross

Hi Ross,

Good to see this being worked on. The Orbital project is certainly interested in this. 

In academia, datacite is being pushed as the defacto standard for data citation. It's handle/DOI based and led by the British Library. There have been workshops recently at the BL and quite a lot of write ups on blogs. Although datacite is a service for minting DOIs for data, it also makes recommendations for citation formats, which CKAN could adhere to, even if CKAN didn't include the option of using datacite for minting of DOIs (which costs money).

Here are some useful links:

List of data archives: http://datacite.org/repolist These aren't all using datacite by any means, but it's interesting to see how they publish research data.

http://mds.datacite.org/static/apidoc API docs!

http://schema.datacite.org for metadata required. 

For Orbital, I'd like to ensure that all the datacite mandatory and optional medata attributes are included by default:

Mandatory (5): http://schema.datacite.org/meta/kernel-2.2/doc/DataCite-MetadataKernel_v2.2.pdf (page 8)
Optional (12): http://schema.datacite.org/meta/kernel-2.2/doc/DataCite-MetadataKernel_v2.2.pdf (page 11)

Some of these are already being captured by a vanilla CKAN install, others we could add ourselves, but it maybe its a useful benchmark for CKAN's default metadata? Page six of that PDF document also gives guidance on Citation formats. They recommend this format:

Creator (PublicationYear): Title. Version. Publisher. ResourceType. Identifier

Example:

Irino, T; Tada, R (2009): Chemical and mineral compositions of sediments from ODP Site 127‐ 797. Geological Institute, University of Tokyo.doi:10.1594/PANGAEA.726855. http://dx.doi.org/10.1594/PANGAEA.726855

Using this format, your example citation would look like:

Song, Steve. (2012) African Terrestrial Fibre Optic Cables. the Data Hub. http://thedatahub.org/dataset/afterfibre

You can use http://search.datacite.org to find data that is using datacite and study the citation format. e.g. http://doi.pangaea.de/10.1594/PANGAEA.57053 orhttp://esds.ac.uk/doi/?sn=86#0

and finally, http://test.datacite.org is a self-contained testing environment if you/we decide to write a datacite extension for CKAN.

Cheers
Joss



--
Joss Winn
Contact: http://lncn.eu/joss
Calendar: http://lncn.eu/az27