If you ask a little more, people are willing to share data, but only under appropriate conditions and control.
At present, many data are not yet openly accessible for educational and research purposes. This is despite the fact that the value of data would increase if others could also use it. However, institutions still experience too many obstacles to sharing data. These can be for legal reasons concerning ownership, intellectual property, authorship, privacy (AVG), ethics, consent (not obtaining permission to use data), competition, or commercial interests. As soon as there is any risk involved, an institution is quick to dismiss it. "If you ask a little more detailed question, however, people are quite willing to share data - they have an interest in it too - but only if it is done under appropriate conditions and a certain level of control", says Axel Berg, Innovation Manager for Labs at SURF. "You see that many institutions have not yet clearly defined the conditions under which they would like to share data. Because that's what it's about: drawing up clear conditions and looking at how you can enforce and monitor them.”
Standards and formats
The most important obstacle to data sharing is lack of control, agrees Ron Augustus, SURF’s CINO (Chief Innovation Officer). "If institutions don't know what happens to their data and what it's used for, they decide not to share it. That's why we're working within SURF on standards regarding metadata and formats. That will make it clear under what conditions you can share data and how you want it to be delivered. The FAIR principle is used to unlock data sets. "Board members in education and research are working to make more and more data Findable, Accessable, Interoperable and Re-usable. In this way, we want to give open science a helping hand."
Various platforms and bodies are now involved. One example is the National Platform Open Science (NPOS), which aims to promote the sharing and opening up of data and in which institutions are represented with the support of the Ministry. At European level, it is the EOSC (European Open Science Cloud). "SURF also has a number of people who are concerned with open science and data sharing. Just like the National Coordination Point Research Data Management (LCRDM), which is coordinated and supported by SURF. And various research communities are also concerned with this, because data sovereignty naturally also plays a role at departmental level," says Augustus.
Putting FAIR on the Agenda
At the institutional level, board members can do their bit by putting making data FAIR on the agenda. "And that not only applies to research data but also to publications, models and scientific software. For all these digital assets, board members should formulate policies to pursue FAIR principles as far as possible and what control they want to have over them," says Augustus. "You can stimulate this by rewarding scientists and departments who make data FAIR accessible. You could further investigate how FAIR your institution is by finding out which departments are already advanced and which are not yet, and by steering accordingly."
To stimulate it, board members can put making data FAIR on the agenda.
When establishing such frameworks and conditions for data sharing, it is good if institutions involve data stewards and researchers from their central research support group or DCC. "Among other things, to see what the technical possibilities are for implementing these control mechanisms. If you promote data sharing and make it easier by enabling it in a trusted and controllable way, you encourage data reuse," says Berg, who has already developed some prototypes around data exchange. "For example, those who want to use the data can do so through a trusted application without being able to see the data themselves. The data owner sees the result first, checks it for sensitivities and then releases it."
According to Augustus, the fact that institutions are increasingly promoting FAIR is also desirable from a social point of view. "We are funded from public resources, so we will also ensure that the output of our institutions is made publicly accessible, including to commercial providers," he says. "My expectation is that data exchange, where data can be shared confidentially under public and/or commercial terms, is going to become a whole industry on top of the internet. Already now, data is the lifeblood of researchers. And soon it will be the lifeblood of commercial companies that want to share data to improve their own processes or services."
To this end, SURF is taking the first steps with the University of Amsterdam in Project RDX: Research Data Exchange. "There we are investigating step by step which use cases require control, what control exactly that is, and how you can implement it. In the meantime, we have developed some interesting additions to existing services that may go into production next year," says Berg. "The implementation of a broad vision of complete data exchange where different groups can share data and set their own conditions - that will still take a few years." (More information about this project can be found on the Amsterdam Economic Board site, among others). The RDX project builds on a vision and extensive current scientific research in the field of data exchanges within the Institute of Computer Science at the UvA, led by Professor Cees de Laat and Professor Leon Gommans.
Control of Data
On 1 October 2021, SURF launched the ‘Regie op Data’ (control of data) project group (as part of the 2-year plan). Encouraging FAIR data is a specific activity within this project group, one of the areas in which SURF aims to accelerate the provision of services. SURF is also involved as a partner in sector initiatives such as Health RI, in which the sharing of data in the specific healthcare context is discussed. Europe is also working on the Data Act, which among other things regulates how you should be able to exchange data. "It's an international theme, and SURF is also working with European parties to see how data sharing can be encouraged and controlled, and what formats and standards are needed. Because if you have those, you can make data exchange easier," says Augustus. "In fact, an institution should not only be judged on the number of publications or the journals in which they appear, but also on how much data they have shared. So that the quality of research improves. That's what SURF is trying to do its bit for."
Tekst: Wilma Schreiber
'Breaking down barriers to data sharing' is an article from SURF Magazine. Monthly the newest articles in your inbox? Then subscribe to the SURF Magazine newsletter (in Dutch).
Questions about this article? Mail to firstname.lastname@example.org.