Home >News> Data-driven experimentation in EUHubs4Data: a report from Data Forum 2023

Data-driven experimentation in EUHubs4Data: a report from Data Forum 2023

21/04/2023

This year’s Data Forum, which Happened in The Hague during the first Data Spaces Symposium, was, as always, rich in content and inputs from the speakers on stage.

Focused on showcasing the accumulated knowledge and experience and presented to an enthusiastic audience seeking to expedite the advancement of data-centric applications and services, the Data Forum has covered many aspects of the project in four sessions.

The first one, “The power of a federation of iSpaces”, offered an overview of the founding pillars of EUH4D: the Federation of over 30 iSpaces, the rich catalogue, the SMEs engaged in the experiments, the ethics framework and methodology for data sharing, and overall visibility of the achievements of the project. The session also stressed the power of the EUH4D Federation of iSpaces: working together is the key to achieving a bigger impact. Therefore, EUH4D results are very well positioned to give value to experimentation and the adoption of data technology thanks to the members’ practical experience.

The second session focused on the challenges and lessons learned in designing, building, and using the federated catalogue for the project. The catalogue is distributed, allowing each partner to manage its dedicated local instance and decide which assets to share with the federation and which to accept from other partners. The project has focused on interoperability with the IDS framework, which many of the hubs have already implemented, to ensure safe and easy data exchange in a trusted environment and manner. The project has also improved the knowledge and skills of iSpaces, which provides local support to companies. On the Dataset side, the need is to finetune the needed workflows to push newly identified datasets to the catalogue and perform automatic pre-analysis together with metadata enrichment to work on metadata harmonisation and datasets classification based on ontologies, and, last but not least, define the policies to share datasets – which come from the hubs, are hosted in external repositories, generated by the funded experiments or from initiatives EUH4D engaged with (e.g. data platforms ICT-13 projects). This is done with an eye on data protection according to guidelines periodically released to partners to guide how to best deal with data. A webinar on the topic will come in late Spring!
Overall, the catalogue has proven its utility to iSpaces and end users thanks to its light technology for great flexibility, user testing, and collaboration among hubs based on the service delivery framework.

EUHubs4Data team at the Data Spaces Symposium

The third session revolved around “Practical Ethics and Compliance for Data (Re-)Use in Innovation Experiments” and why FAIR is not always enough, with first-hand experience from different players in the project. One of the experiments funded by the EUH4D Open Call series participated with the lessons learnt from reusing publicly available data for data scraping, for both copyright and privacy reasons. Lastly, there was a useful overview of the lessons learned on data reuse from a Data Management Plan perspective: to make an example, the template produced for the experiments to fill in the Data Management Plan that ended up being subject to different interpretations of the requirements or of the actor making datasets FAIR before being changed for the third open call. Overall, the DMPs analysed lacked complete information because of little knowledge of the DMP as an instrument to monitor, not sanction.

The concluding session of the Data Forum, “Data Act and Data Governance Act: hurdles or opportunities?”, offered a thorough but concise overview of the barriers for start-ups and companies wishing to use AI and Big Data according to the provisions of the Data Governance Act, in force and applicable from September 2023, the concept of neutrality of the data intermediary and the possible effects on the Federation. In fact, despite the lack of definitive answers, something can already be agreed on, like that the Federation will possibly qualify as a data intermediary as well as the individual iSpaces, and that there will be associated costs to comply with the DGA. However, some advantages could also be identified, such as that data intermediaries are distinct from the typical for-profit actors and that their neutral actor position can be leveraged to secure easier, cheaper or more affordable access to public sector data. The presentation also touched on the Data Act, its different versions and the potential inherent ambiguities. To conclude with some optimism, despite a load of requirements, adhering to the European Common Data Spaces harmonised standards should imply that those requirements are met. Each data space’s participants may assign responsibilities for the requirements between themselves, thus reducing the complexity.

We are looking forward to seeing what the developments of these two acts entail for the Federation.

Don’t miss participating in DataWeek 2023!