Open for all UnRAVeL Members: Talk: Andrzej Wasowski

Wednesday, May 3, 2023, 4pm

Location: RWTH Aachen University, Department of Computer Science - Ahornstr. 55, building E3, room 9222

Speaker: Andrzej Wasowski



Disclosure of data analytics results has important scientific and commercial justifications. However, before you share any data, you must understand what is the risk that the it carries private information about the subjects. Probabilistic programming and information theory can help to explore information leakage in data analytics programs. In this lecture, I will explain the basics of what information leakage is, and demonstrate it using examples from research on differential privacy, genetics, wine tasting (!) and general population statistics.  The examples will be analyzed using our privacy debugger, a Privugger.

While I may touch on the formal semantics of probabilistic programs, most of the talk will be accessible for researchers generally knowledgeable in data analysis.

The presentation will also briefly introduce the IT University of Copenhagen (ITU), one of the very few higher education and research institutions world-wide, devoted solely to Information Technology.




Andrzej Wasowski is a professor of Software Engineering at IT University in Copenhagen. He works on methods, tools, and architectures for systematic design of trustworthy high-quality software systems. His goal is to expand the innovation capabilities from narrow groups of experts in academia to engineers and data analysts. His research includes understanding how intelligent systems fail, what problems programmers face building them, what questions they ask about the systems, and how information is manipulated by these systems. Wasowski has collaborated with open source projects (Linux kernel and ROS among others) and with industry (Danfoss, Eiva, Bosch). Currently, he is investigating quality assurance methods for robotics platforms, the privacy and information flow in machine learning programs, and generating patches for locking bugs in the Linux kernel.