EOSC-Nordic at the RDA machine-actionable Data Management Plans (maDMPs) hackathon


At the end of May, EOSC-Nordic members Hanne Moa (UNINETT, Norway), Joakim Philipson (Univ of Stockholm, Sweden), Anne Fouilloux (UiO, Norway), Adil Hasan (UNINETT Sigma2, Norway) took part in the RDA machine-actionable Data Management Plans (maDMPs) hackathon to bootstrap adoption of the maDMP schema in the Nordics.

The machine-actionable DMP working group has produced a first version of the schema for machine-actionable data management plans, and there have been several efforts to implement the schema in various data management tools. The goal of the hackathon was to collect together these different efforts such that those working on the same aspects can share ideas and approaches. The hackathon was quite global, with approximately 40 people from the USA, Australia, Europe, and South Africa attending.

One of the tasks in work package 5 (T5.3.2) is to integrate the data management plan into the data management process. One way to achieve this is by making the plan machine-actionable such that steps in the process can be executed (for example, reserving storage space). Or, the information in the plan can be harvested from data management services (for example, DOIs for datasets). In this way, the plan is more realistic as it drives or is driven by the process. The machine actionable DMP schema produced by the RDA working group is very useful as a means of communicating between the DMP tool and the services that make up the data management process for a project. The RDA is an international organisation, and by adopting an RDA approved schema, we are more likely to be interoperable with tools and services managed in other countries.

The EOSC-Nordic group focussed on three areas of the hackathon: adoption of the maDMP schema by DMP tools, mapping DMP templates to maDMP schema, and mapping the maDMP schema to executable workflows. Our goal was to gain sufficient knowledge and collaboration to enable our tools to export a plan in the maDMP schema and import a plan conforming to the maDMP schema. Another goal was to be able to create a plan template that conforms to the schema and to be able to include parts of the data management process into the maDMP schema. We didn’t expect to complete any of these within the time of the hackathon, but we planned to make connections and have established how to achieve these goals.

We managed to make connections with groups working along similar lines and were able to export a DMP from easyDMP and have it imported into another DMP tool (DMPOnline), demonstrating the interoperability of DMP tools and made some suggestions for the maDMP schema to achieve our goals. We also attracted interest from some groups in the work we are doing in integrating workflows in the maDMP. So, we achieved what we wanted and will continue with the work.
Going forward, we will continue to take part in the maDMP meetings and report our progress to the working group. We will also look to exchange our progress with the members that expressed interest.