Use Case: Brokered data (also could be used for other data services)

[rduerr]

• Contributors: Ruth Duerr & the BCube team

Specify if this is a "Citing Software" and/or "Citing Data" Use Case: Both

Goals and Summary

How would middleware (e.g., GI-Cat/GI-Axe) which mediates between data services provided by a repository (including things like data sets available via ftp, OPeNDAP, OGC W*S, etc. or advertised through OAI-PMH, THREDDS, web-enabled folders, etc.) provide users of its services, citation information? What would the citation(s) cover? Who would they credit? In what roles?

Assuming that the RDA-devised strategy for dealing with the subset specification (i.e., dynamic data) were implemented by repositories, how would a broker or other such middleware service generate a citation to pass along to whatever client the user is using (e.g., arcGIS, IDL, R, Matlab, project-specific GUI, etc.) to support this form of reproducibility?

Why is it important and to whom?

• Reasons

Multirepository data discovery and access interfaces and services are becoming widespread. In many cases, they involve a chain of web services on subsets of data from multiple places. It is not at all clear how/who to credit in these cases where the user may not even really be aware of where the data is coming from or who created it or how it was created.

Having a reasonable citation method in these cases would help the science enterprise be more transparent, give credit where credit is due, and also support reproducibility. In other words, the folks who should care are funders, data creators, repositories, tool developers, publishers and the like.

Why hasn’t it been solved yet?

• First reason

Regular data citation practices are not yet the norm and those are the simple cases.

• Second reason

Most repositories do not yet support the dynamic data citation method proposed by RDA, so have no starting point to work with.

• Third reason

It's a hard sociotechnical problem

Actionable Outcomes

Workshop outcomes will include actionable plans to enable the broader research community to implement the software and data attribution practices that are identified and advanced by the participants of the workshop. Articulate plans to implement this use case, as possible, into one or more of your projects subsequent to the workshop.

If a viable solution were proposed the BCube project would work to implement it in GI-Cat/Axe.
Also, NSIDC would consider implementing the parts in its control as funding permitted.

Additional Information and Links

• http://nsidc.org/research/informatics/bcube