Machine-Readability Facilitates DMP Efforts

You are here

Machine-Readability Facilitates DMP Efforts
26 May 2020 UTC

Machine-Readability Facilitates DMP Efforts

26 May 2020

The purpose of a data management plan is to describe the management of data in a research project throughout the research process, in order to provide various stakeholders with the right information when they need it. There are currently several different DMP templates around the world, and a great need for coordination. A global initiative for creating a common, machine-readable – or machine-actionable – standard for data management plans is the RDA working group DMP Common Standards Working Group. In a recent RDA Sweden and SND webinar on the development of machine-actionable data management plans, the results from the working group were introduced, as well as some examples of the development work in Swedish higher education institutions.

A driving reason for creating a global, machine-readable standard for data management plans is that it would facilitate automatic information exchange between tools and systems used in research. This could lead to a decrease in the work efforts and administration which are often associated with DMPs.

Today it's difficult to get an overview

– Many of the existing templates and systems for DMPs have shortcomings. They have to be manually completed, they may be vague in which information they are asking for, and they’re often considered bureaucratic by researchers. It’s difficult for an individual researcher to get an overview of the existence of these tools, and to know which specific tool to use, said Tomasz Miksa from the RDA DMP working group.

An important focus for the working group has been to describe an automated workflow in the creation of a DMP. This means that the contents of the DMP aren't generated just by the researchers and the systems they use, but also by the infrastructure support for data management in for example an HEI (see the graphic). A workflow like this would decrease the workload for the individual researcher, as well as the risk that outdated or incorrect information is entered into the DMP.

The working group has also analysed which types of information that a machine-actionable DMP may contain. Their analysis shows that an automated Data Management Plan makes it possible to gather more as well as more detailed information than traditional, manually created plans. (You can read more about what kind of information this may involve in the recommendations of the working group.)

Experiences from Swedish HEIs

There are several Swedish initiatives for developing templates and digital tools for data management plans, such as a national working group chaired by the Swedish Research Council. The plan is to have a digital DMP tool available during 2020. Participants in the RDA/SND webinar were given introductions to some of the digital DMP tools which are used locally in the HEIs. They are often based on international tools such as DMP Online (Stockholm University), Data Stewardship Wizard (SciLifeLab at Karolinska Institutet, KTH, Stockholm University, and Uppsala University), and DMP Roadmap (Lund University).

Just like in the international setting, the discussions in the national development efforts are focused on how to make DMP work simpler, less time-consuming, and more automated. Joakim Philipson from Stockholm University highlighted some possible solutions, such as tools with more drop-down menus and pre-populated response alternatives in order to make the information machine-readable, and to use established metadata standards to a greater extent than today. Maria Johnsson from Lund University emphasised the importance of getting researchers involved in the development efforts right from the start:

– The central person is the researcher. A tool has to be attractive, and we must have continuous discussions with researchers about this. The system we use in Lund is very simple, the routines are easy to implement, and we have put a lot of focus on user support.

The webinar on machine-actionable data management plans will be followed up after summer within the SND operations. If you want to be able to influence the developments and recommendations on a national as well as international level, you are welcome to take part in one of the RDA interest or working groups:

Active Data Management Plans Interest Group
Exposing Data Management Plans Working Group
DMP Common Standards Working Group

Here you can have a look at the presentations and a video recording from the webinar.