Privacy-Friendly Data Analytics
User data are constantly collected with various organizations for the purpose of aggregate analysis. Here, the analysist is not interested in individual user data, but rather in user data in the aggregate. Nevertheless, the privacy of individual user data is at risk, for two fundamental reasons: the query result may leak too much information, or the data aggregator itself may leak collected data, intentionally (e.g., selling) or unintentionally (being comprised). A further complication comes from the fact that the analyst is often interested in querying the dataset obtained by joining data collected by different organizations, which poses a secure data sharing problem. In other words, the privacy of data analytics is a two-fold problem, which encompasses the sanitisation of query results as well as the secure storage and sharing of user data.
The project consisted of five work packages in order to solve the above mentioned problems and realize the overall aim to design an effective and privacy-preserving architecture for data analytics.
Role Within the Collaborative Research Center