Building blocks

Contact

Pepijn Groen

Coordinator working group Data Sharing

Data Sharing

AI machine learning is impossible without data. The greater the amount of relevant data available, the better the predictive value is and so the more useful machine learning AI applications become. This means that access to data is crucial. In the Netherlands, data is often kept locked away, mostly for legal or commercial reasons. To break down those barriers, data sharing needs to be organised properly and responsibly, much better and more rapidly than what we’re accustomed to right now. More access to data means faster AI implementation and greater accuracy, resulting in better service. This is all based on confidence, understandings, more knowledge and our democratic principles.

Data availability is an essential success factor in AI development. The greater the amount of relevant data available, the better the predictive value of an algorithm gets. And that improved predictive capability in turn leads to improved AI solutions. Access to data, both within organisational boundaries and across them, is crucial for using AI successfully. Data is often kept stored within the boundaries of organisations and shielded from the outside world. Technical, legal and commercial restrictions and interests stop organisations from making their data readily available to others. That’s a pity, as massive social and economic opportunities can appear when organisations from the commercial sector, sciences and the public domain can use each other’s data for developing AI applications. The Data Sharing working group is committed to creating a flourishing data economy in the Netherlands that will make numerous new and improved AI applications possible and accelerate the implementation of AI.

Working group approach

Three substantive topics are important for creating that thriving data economy:

  • Dataspaces: ecosystems in which access to data from other organisations is organised through common agreements.
  • Semantics: to make sure that the organisations within a dataspace understand each other’s data and algorithms so that cooperation is possible.
  • Privacy-enhancing technologies: allow data to be shared in cases where privacy and security are a major challenge.

As a working group, we therefore have three main goals:

  • Setting up dataspaces We encourage the development of ecosystems in which organisations exchange data for the benefit of AI applications. The working group supports various organisations in launching ecosystems that focus on data sharing for AI applications, which in turn helps to demonstrate the value of sharing not only data but also the practical lessons. Offering tools, expertise and guidance and helping to identify possible funding opportunities lets the working group help these organisations to accelerate.
  • Developing a trust system We are strongly committed to standardised, scalable instruments and tools that make it easier for market parties to set up dataspaces. To help organisations launch dataspaces on a large scale, the Data Sharing working group is working on developing standardised and scalable tools that will make the launch of a dataspace easier. This also lets us ensure harmonisation across dataspaces: if different dataspaces use similar architectures, it becomes easier to create links between them.
  • Activating the market We largely do this by actively disseminating knowledge about how valuable data sharing is for AI. To realise the objectives of the working group, market parties need to be made aware of the value of data sharing for AI and the possibilities it creates. That is why the working group actively disseminates knowledge about the subject. As well as organising knowledge sessions and publishing reports, the working group will in future be launching innovation centres that help businesses share data for AI applications.

To coordinate all the developments above, we maintain close contacts with the leading initiatives through our network. We do that at the European level (for example Gaia-X and IDSA), the national level (for example the Data Sharing Coalition and FAIR) and within specific dataspaces (iSHARE and SCSN).

More information

If you have an idea for a data-sharing AI application, please take a look at this publication. You can find use cases here if you’re looking for inspiration. If you are interested in taking part in the Data Sharing working group, please contact us to see what we in the working group could mean for you.

Check out related data sharing downloads

Share with