scientific data journal - data citation example

data-citation.xml

<ref-list content-type="data-citations">
<ref id="d1"><element-citation>
<named-content content-type="source">existing</named-content>
<source>Dryad Digital Repository</source>
<ext-link ext-link-type="dryad" specific-use="id" xlink:href="dryad.rv624">dryad.rv624</ext-link>
<year>2010</year>
<collab><contrib-group>
  <contrib>
    <name><surname>Brown</surname><given-names>Alison</given-names></name>
  </contrib>
  <contrib>
    <contrib-id contrib-id-type="orcid">http://orcid.org/0000-0002-6199-2343</contrib-id>
    <name><surname>Xiang</surname><given-names>Xu</given-names><suffix>Jr</suffix></name>
  </contrib>
</contrib-group>
</collab>
</element-citation></ref>
<ref id="d2"><element-citation>
<named-content content-type="source">new</named-content>
<source>SmallRepo</source>
<ext-link ext-link-type="smallrepo" specific-use="url" xlink:href="http://smallrepo.org/data/787700">787700</ext-link>
<year>2004</year>
<collab><contrib-group>
  <contrib>
    <contrib-id contrib-id-type="orcid">http://orcid.org/0000-0002-8205-121X</contrib-id>
    <name><surname>Smith</surname><given-names>Mary</given-names></name>
  </contrib>
  <contrib>
    <name><surname>Jones</surname><given-names>Peter</given-names><suffix>Jr.</suffix></name>
  </contrib>
  <contrib>
    <contrib-id contrib-id-type="orcid">http://orcid.org/0000-0002-5463-0974</contrib-id>
    <name><surname>Duncan</surname><given-names>Andrew</given-names><suffix>III</suffix></name>
  </contrib>
</contrib-group>
</collab>
<etal/>
</element-citation></ref>
</ref-list>
 

this is an example of data citation as provided by the Nature journal Scientific Data, and it was described in a message from Paul Donohoe on 2014-05-02 on the JATS mailing list.

Paul describes the tagging as follows

I've used the existing JATS Journal Publishing tag set. We decided to use as we want to list them like article references and have the same styling, and it caters for our requirements. Aside from the obvious elements, we also wanted to distinguish between data citations created as part of the work described in the article ("new") and data citations referenced by the article ("existing") [ I'm not completely happy with these labels ]. We use a element for these. also allows use of where we may capture only the first few authors of a data citation.

One of the challenges we thought about was if/when data repositories arbitrarily change the way they determine citation URLs (we did have one example of this a few years ago). Rather than store the absolute URL in the article XML, we decided to create an ontology of URI templates, giving each template an identifier, and we put the identifier in the XML. We then look up the identifier in the ontology when rendering to HTML. Then we can easily and quickly change the URI template if a repository does change its URL structure.

We also decided to cater for where a repository does not have a fixed URL template (we haven't seen any of those yet). In that case, we do store the absolute URL in the XML, and know that it could break if the repository changed.

The sample XML below assumes that Dryad does use a URI template, and the fictional SmallRepo does not.

Was there a reason for not using the full DOI in your Dryad example above?