Text by: Dr. Jeff Love, Stockholm University Library
A couple of weeks ago the conference ‘Putting the EOSC Vision into Practice’ in Porto, Portugal. It was arranged by EUDAT, an international organization funded by the Horizon 2020 scheme dedicated to developing infrastructure for data management. As the conference title implies, the sessions here were intended to demonstrate practical aspects of the forthcoming European Open Science Cloud (EOSC). It will come as little surprise that only the preliminary questions were posed here: how to organize all of the people and services and, perhaps more importantly, what the EOSC will actually do.
Right now the prevailing thought seems to be that the EOSC will be a federation of existing infrastructures: repositories, computing centres and other Europe-based organizations connected to research data. Naturally this definition is extremely broad, so it remains unclear how much the EOSC will include, as well as how it will operate. Will it be a collected portal where researchers and companies can search for specific services? Will it offer consultancy services for training? Will it serve as an accreditation body for open data services? I can’t help escape seeing a conflict-of-interest in having such an endeavour being run by an organization with its own active data services, though perhaps this is unavoidable. Indeed the EUDAT services were heavily marketed at the conference (see the link for a list and descriptions).
A panel of experts was asked what they wanted out of the EOSC, and the responses from Alex Vermeulen (ICOS) and Françoise Genova (Strasbourg Astronomical Data Centre) were the most telling: the EOSC should be a 100% efficient, almost invisible system with reliable services matching researcher user needs, including full integration of services and frameworks currently in use. Alex also pointed out that ‘researcher user needs’, while often mentioned, remain poorly understood.
All of these questions remain open, and there are at present several task forces assigned to assess them and make recommendations. One major development is the start of the EOSC-Hub, which seeks to provide a searchable catalogue of existing data services. The EOSC-Hub has 100 partners and will be constructed between Jan 2018 and Dec 2020. The project leader is Per Öster from CSC in Finland.
In addition to panels on the EOSC, there were a number of breakout sessions dealing with matters such as research library involvement in data management, the handling of sensitive data and data training. One session highlight was David Foster’s lively presentation on the implications the imminent General Data Protection Regulation will have at CERN. If you meet him, ask for his catchy jingle about it. Heiko Tjalsma discussed plans for the adoption and adaptation of Harvard’s Data Tags for use at DANS. This may prove a simple, clear solution for managing sensitive data at other repositories as well.
Powerpoint slides from several presentations have been made available through B2DROP and can be accessed via the conference programme.
Want to get involved?
Individuals (researchers, data managers, librarians, anyone really) can be a part of research interest and working groups, such as those listed at the RDA. Organizations and service providers which want to partner up with EUDAT can look into doing so here.
If you want to know more about Stockholm University’s policy on Research data and Open Science, please visit Stockholm University Library’s website.