Over the past few months, we've been working on several improvements to the support that RSpace has for Data Management Plans (DMPs). This has included adding the capability to import DMPs into RSpace from Argos, integrating with Zenodo in such a way that allows for the export of data from RSpace whilst referencing such a DMP, and finally by working towards becoming a participant in the Research Data Alliance's effort to standardise machine actionable DMPs, for the purposes of interoperability. It is this third piece of work that is not necessarily front-and-centre in the product, and we thought it worth explaining what we've done and why.

One of the most important aspects of the RSpace product is to be the interocular with many of the tools that researchers could want to use as part of their work.  The ecosystem of tools RSpace supports is always growing. The current ecosystem is depicted in this graphic

Rather than dictate what tool to use for document storage, or collaboration, or publishing we aim to be flexible to the tools researchers already use; linking together data and services that are otherwise siloed. One of the best examples of this is our export-to-repository functionality where we support exports to Dataverse, Dryad, Figshare, and Zenodo which allows researchers to share their experimental data recorded in RSpace in whichever way works best for them and their organisation. Most importantly, we've developed these integrations in such a way that allows us to scale the maintenance burden of supporting all these difference services efficiently so that we may continue to be able to easily support more into the future.

Simply supporting lots of different publishing services is not quite enough though: an ad-hoc approach to the publishing of experimental data complicates the traceability and integrity of that data. Machine-actionable DMPs aim to solve much of this issue by allowing for data to be described in a uniform way and then referenced across systems wherever the data may flow. The Research Data Alliance (RDA)'s DMP Common Standard is an attempt to standardise the structure of machine actionable DMPs so that they may be transferred between systems in an interoperable way. Therefore, it makes complete sense for RSpace to be compliant with this specification, so that we may support any and all tools that researchers may wish to use that provide or consume DMPs in a compliant way.

Much of the work completed thus far to make the RSpace product compliant with the specification has been completely behind-the-scenes adjustments to how we import DMPs from DMPTool, with no immediate changes to the Web interface. For example, the specification describes ancillary information about the data that we can ingest (like projects and funders), but also key information like a DMP’s DOI or the ORCID IDs of the participants; data that help link together various information systems that will only be more important as we continue to focus on interoperability and information identifiers. To extract this data, we have encoded the data model as described by the specification into Java and published that code under an open-source license to GitHub.

Currently, we are using this data model as the basis for importing DMPs into RSpace from services that provide compliant DMPs via their API (Application Programming Interface), like DMPTool. It will also be the basis for future work where we can support enriched data flows through other compliant tooling; referencing a DMP as part of an export to a repository is just the first of what we envision will be various common tasks performed inside RSpace that will be enriched by using machine actionable DMPs.

‍