Start:
2021/06/10
End:
2021/06/10
Location:
Online
Link:
4th SoBigData++ Awareness Panel
Description:
15:00 - 15:15: Studying population dynamics without compromising people's privacy

Speaker: René Peralta (NIST, USA)

Abstract: We discuss the measurement of aggregate levels of encounters in a population, a concept we call encounter metrics. The technique is designed so that it can be deployed while preserving the privacy of individuals. We compare the privacy profile of encounter metrics to that of GAEN, the Google/Apple exposure notification system.

15:15 - 15:30: Data Sovereignty in the connected vehicle

Speaker: Frederick Richter (Foundation for Data Protection, Germany)

Abstract: When we drive individually, we can perceive this as experienced freedom. But increased collection and processing of mobility data could make the moving data subject an object of surveillance. Which technical or legal measures can assure natural persons' data sovereignty? Which extent of transparency and control has to be offered to the mobile persons to secure their citizen rights - and what can be effective instruements to offer it? The talk will highlight aspects of anonymity, of trust building mechanisms such as privacy dashboards and of potential modifications of the GDPR.

15:30 - 15:45: The MobiDataLab project

Speaker: Thierry Chevallier (AKKA Technologies, France)

Abstract: More and more data are produced in the transport sector, with new mobility services appearing almost every day. Mobility stakeholders agree that data sharing can unlock new insights and lead to more efficient processes and new products. However there is still a lot of reluctance to share data, for very different reasons. Among these barriers to data sharing, trust issues and privacy concerns are perhaps the most important. In order to address this problem, the MobiDataLab project, funded by the EU, works to foster the sharing of mobility data amongst transport authorities, operators and other mobility stakeholders in Europe and beyond. The use of mobility and location-based data can raise justified privacy and ethical concerns, that is why MobiDataLab is developing knowledge on this topic as well as a cloud-based solution to make data sharing more secure. This presentation aims to describe the prototype under development, which uses privacy preserving and anonymization techniques in order to safely share mobility data. This Transport cloud will be validated during Living Labs addressing local specificities from transport authorities regarding, among other challenges, data privacy.

15:45 - 16:00: Decentralized anonymization of mobility data

Speaker: Josep Domingo-Ferrer (Universitat Rovira i Virgili, Catalonia)

Abstract: Collecting trajectories is of paramount importance to improve public health, transportation, urban planning, economic planning, etc. However, trajectories are personally identifiable information and should be anonymized before publishing or sharing them for secondary analysis. The standard approach to build anonymized sets of trajectories is centralized: the subjects send their original movement data to a controller, who takes care of producing an anonymized mobility data set. A step further towards privacy is to empower the subjects with the ability to anonymize their trajectories locally. The implications of decentralized anonymization of trajectory data will be briefly discussed.

16:00 - 16:15: Mobility data reidentification opportunities and… risks

Speaker: Giovanni Comandè (Scuola Superiore Sant’Anna, Italy)

Abstract: Mobility data include a set of data types with different origins and sources but that alone, or combined, give information on how an individual moves, where she usually goes and what activities an individual carries out. We discuss their enormous potentials for many uses along with risks. From a legal point of view, mobility data are not considered as such per se sensitive data (as health or political opinions data are). However, what we highlight is how apparently unproblematic mobility data can become risky for privacy when they are combined or thoroughly analyzed with the relevant methods and/or external data. Indeed, even if they are not -per se sensitive personal data they may easily reveal sensitive and confidential information which need to be shielded without preventing proper exploitation.