Mireille Hildebrandt on Data Driven Intelligence and Data Protection Law at the International Conference of Data Protection & Privacy Commissioners (ICDPPC)
October 19th, 2016
Mireille Hildebrandt was one of the panelist of a closed session at the International Conference of Data
Protection & Privacy Commissioners conference taking place in Marrakech this week. Her slides are available
on their public site.
She used the concept of choice architecture, which derives from nudge theory and behavioural economics, to clarify why we need purpose limitation, highlighting the concepts of bias, opacity and pattern obesitas in
machine learning. Explaining how purposes limitation will indeed save us from data obesitas (think twice
before selecting your training data) and pattern obesitas (garbage in, garbage out, algorithms will always see
patterns but not necessarily those that are relevant).
PEP Project Receives €760 000 Funding
October 10th, 2016
The PEP project builds on the Polymorphic Encryption and Pseudonymisation technique developed by Bart
Jacobs and Eric Verheul. This technique stores pseudonymised data in encrypted form. Researchers working on the data can only decrypt the parts of the data for which they have access rights.
See http://pep.cs.ru.nl/ for more details.
The collection and analysis of medical data on a big data scale is becoming an essential approach to
understand complicated diseases. In order to gain new insights it is important that international researchers
can cooperate: they will have to access each other’s data and contribute to the data sets. In many cases, such medical research involves privacy sensitive data about patients. Patients should be able to count on
preservation of their privacy and on secure storage of their data. All this should comply with upcoming
European privacy regulations, which are the most stringent in the world.
The PEP project has its own funding from public sources, with a current budget of €1.6M; €760K is funded by the Province of Gelderland, and the remainder by Radboud University. Eventually the software will be made publicly available under an open source license.
Tweemaal het PI.lab in De Correspondent
Onlangs verschenen in de Correspondent twee artikelen over PI.lab onderzoekers: Corien Prins (over de
risico’s van Big Data; Corien is lid van onze bestuursraad) en Bert Jaap Koops (over zijn Vici onderzoek naar een nieuw paradigma voor privacy bescherming).
Interview Mireille Hildebrandt in De Correspondent
Onlangs verscheen een interview in De Correspondent met PI.lab hoogleraar Mireille Hildebrandt, over het belang van transparantie, toezicht en handhaving bij data-analyse toepassingen.