Topics for thesis
If you want to do the thesis with us, please kindly contact the researcher in charge of the research field you are interested in:
Patricia Arias-Cabarcos works at the intersection of human computer interaction and security. Her research covers new authentication mechanisms based on behavioral data and usable security & privacy in general.
Simon Hanisch current research interests are in the privacy of humans in motion and network security for edge clouds: Our motions, such as our walking pattern, can be used to identify us and to learn private information such as our gender, weight, or medical conditions. Now with the rise of augmented reality and virtual reality devices, our motion data becomes easier to record and more accessible. For this reason, Simon Hanisch is developing new anonymization techniques for human motions. We work with anonymization techniques, machine learning, motion capture suits, and mixed reality goggles.Topics include the anonymization of human gait and gestures. Edge clouds are a new cloud paradigm that reduces the latency for applications by moving them to the network edge. This new position creates new challenges regarding the availability of applications and the confidentiality and integrity of network traffic.
Christiane Kuhn is interested in understanding the underlying principles of anonymous communication networks and related privacy problems to allow for formal analysis and verification. Anonymous communication networks hide communication meta data, e.g. who is visiting the website with information for cancer patients, as this information would allow an adversary to conclude private information, e.g. that somebody likely is ill. Topics supervised by her focus on formal privacy analysis of existing proposals for anonymous communication networks, improvement of analyzing tools and the investigation of underlying principles, like efficiency limits or basic building blocks.
Christoph Coijanovic is (also) interested in the principles of anonymous communication. His focus lies on anonymous group communication. Compared to one-to-one communication, an additional challenge is the management of groups, i.e., providing mechanisms for users to join and leave groups in an anonymous way. Further, group communication can reveal unique metadata to the adversary, for example, how many users receive a given message or which group a message is send to. Identifying the receiving group might reveal valuable context of the communication, e.g., if a users sends a message to an Alcoholics Anonymous group, the adversary might assume that this user is a recovering alcoholic. Topics supervised by Christoph Coijanovic include the formal analysis of existing group communication protocols, the extension of one-to-one protocols and theory of anonymous group communication.
Javier Parra-Arnau applies mathematical formalisms from information theory, data compression and convex optimization to design protection mechanisms that can prevent profiling, classification and information leakage in the online world. His research is concerned with the pieces of personal data that people voluntarily, and involuntarily, leave on the Internet (e.g., search queries, tags, “likes”, movie ratings and web pages visited), a combination of which may potentially reveal sensitive personal information such as sexual and political preferences, religious beliefs and health-related issues. Part of those data are necessary to enable personalized services (e.g., Amazon’s product recommendations and Google’ search results), but pose serious privacy and security concerns. Javier Parra-Arnau is happy to supervise thesis topics within the following more specific research areas: Data anonymization, User profiling or Privacy quantification.