property for identifying the repository

[timothy...@ec.europa.eu]

Hi,

 

For the purposes of integrating with the Thomson Reuters DCI, in the DataCite metadata I need to distinguish between publisher and repository.  The definition in the DataCite schema documentation for Publisher is very clear, reading 'The name of the entity that holds, archives, publishes prints, distributes, releases, issues, or produces the resource. This property will be used to formulate the citation, so consider the prominence of the role.'.  This definition though says nothing about the actual repository, in which I would be grateful to know whether a property exists for specifying the repository (that preferably supports an attribute for identifying the URI of the repository) ?

 

On a related matter, I wonder why the publisher and not the repository is recommended as part of the pattern for the citation ?  If I draw a parallel with traditional publications, Springer would be the publisher and Journal of Materials Engineering would equate to the repository.  In said example though, the journal is used to formulate the citation, not the publisher.

 

Advice on either topic would be much appreciated.

 

Thanks, Tim

[Joan Starr]

Thank you for raising this issue. I'm afraid I would disagree with you slightly in that I think the description that is currently in place is somewhat less than clear, as it allows for the placement of either a repository (an entity that holds or archives an object) or a publisher (an entity that publishes, prints, distributes, releases, issues, or produces an object). The reason for this ambiguity is that the DataCite Metadata Schema is intended to serve a very broad spectrum of researchers working with all kinds of research objects.

I am very familiar, however, with the particular challenge you are describing in connection with the interoperability with the Data Citation Index. Equally, inasmuch as DataCite is an endorser of the Jt Data Citation Principles (https://www.force11.org/datacitation) and they show repository as part of the citation rather than publisher, perhaps it is time for us to take another look at our recommendation.

In any case, you've given us a lot to think about, and I will add this suggestion to our list for discussion.

Regards,
Joan Starr
Chair, Metadata Working Group

[Doug Cooper]

FWIW in response to Tim's question, I interpret Publisher in the
most traditional manner applicable (i.e. giving priority to the
print publisher), and use:
<contributor contributorType='HostingInstitution'>
<contributorName>...</contributorName>...
</contributor>
to provide the repository name, which I assume can be a URL. Does
anybody handle this in some other way?

Doug

> --
> You received this message because you are subscribed to the Google Groups
> "DataCite Metadata" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to datacite-metad...@googlegroups.com
> <mailto:datacite-metad...@googlegroups.com>.
> For more options, visit https://groups.google.com/d/optout.

[timothy...@ec.europa.eu]

In reply to Joan, my understanding of a publisher is an *organization* that hosts and operates a publishing infrastructure, while a repository is a specific *platform* where objects of interest reside.  In which case, a publisher and a repository are sufficiently distinct concepts to justify their being distinguished one from the other.  While I understand that the present ambiguity offers some added value, I am not sure it justifies precluding the possibility to identify the repository and the publisher separately.  Could Publisher perhaps be extended with repository and repositoryURL attributes ?  Rgds, Tim

[timothy...@ec.europa.eu]

In reply to Doug, are you perhaps using this construct though because there is nothing else suitable i.e. a property or attribute that is clearly intended to identify the repository ?  I ask simply because I would associate 'HostingInstitution' with an organization rather than a repository.  Rgds, Tim

[Joan Starr]

Hello Tim,

The Metadata Working Group has had an exchange with Thomson Reuters on this topic, and here is what TR stated: "In relation to the question at hand, we feel that provision of metadata and inclusion in the DCI is not dependent on the provider distinguishing the source as a repository or publisher in the XML. So long as we as editors are happy that the resource (and metadata) fall within scope it can be included. From experience I would think that application of an additional attribute would be difficult for some providers. In any event, we always strive to seek clarification if anything relating to the selection/metadata is unclear, but we have the scope to include repositories, or publishers. We do ask that the DOI provides a link to the landing page for the data themselves, and not another catalog record, and that the metadata record provided adheres as closely as possible to the Force 11 Declaration on Data Citation."

In virtue of this position, and because we are interested in keeping the schema as simple as possible, we have decided not to add any complexity to the Publisher property at this time.

Thank you for engaging with us in this matter.
--Joan Starr
co-chair of the Metadata Working Group