The York-Maastricht strategic partnership - a data science update
Wednesday, March, 18 was supposed to be the kick-off day for the York-Maastricht partnership announced earlier in the year. Set around five different themes (agrifood, nutrition and health, Europe, global South, imaging and data science), the partnership was supposed to bring together, for the first time, research and education experts from the University of York and Maastricht University (particularly the Institute of Data Science and Maastricht Working on Europe), in an attempt to establish joint research projects, teaching collaborations, knowledge exchange, student and staff exchange and sharing of best practices around university activities.
Unfortunately, that meeting never took place, because of the onset of a global pandemic of unprecedented proportions. However, where there’s will, there will be a way. In the data science theme, efforts to bridge participants from the two universities continued with numerous Skype calls. The full team brings together Dimitris Kolovos, Delaram Kahrobaei, Alfa Yohannis, Nicholas Matragkas and Alfonso De La Vega on the one hand of the English channel (York), as well as Michel Dumontier, Catalina Goanta, Thales Bertaglia, Visara Urovi, Vikas Jaiman, Birgit Wouters, Chang Sun, Mathieu Segers, Jerry Spanakis, Giulia Piccillo and David Townend on the other (UM). The extended team, covering a vast array of multidisciplinary expertise, ranging from data science and history to law, ethics and macroeconomics, will collaborate on various aspects of responsible data science by design.
Today’s technology fails to uphold expectations on how public and private data should be managed and used, mostly because of the centralization of information in the hands of a few powerful players. These disappointments are exemplified by the high profile Facebook/Cambridge Analytica scandal and innumerable abuses of consumer data. The immediate response has been to increase oversight and penalties regarding business practices, but perhaps the fundamental issue lies in the way data is governed and shared for analysis; shouldn’t we, as the audience of these platforms, be able to fully control the availability of our data with other partners? Achieving this in practice requires fundamental reconsideration of design principles for data intensive systems, as well as new data persistence and processing technologies to address privacy, governance, and accountability of what, how, when and why data is analysed. As the foundation of a long-term York-Maastricht research and education program, this vision is unique and compelling because it aims to seamlessly embed the principles of responsible data science (including the FAIR principles), into the actual software development practices used by application developers, rather than simply establishing recommended practices as yet another checklist of requirements or legal/ethical expectations that developers may or may not follow.
So far, the team has been having monthly meetings, in addition to the meetings of the project management team which consists of Michel Dumontier, Catalina Goanta and Dimitris Kolovos. In addition, a Github organization and a Slack work space were created, and are open to other curious researchers who can reach out if interested in the project. Among more palpable results of this interdisciplinary collaboration are an open-source framework for developing secure data vaults (Vaultage) repository, an upcoming research talk, a working paper, as well as a policy brief aiming to create a transatlantic regulatory dialogue with the US Federal Trade Commission. Up next: a hackathon, and a series of lectures on law for computer scientists.