Postings for Theses and Paid Student Positions

Here you can find postings for theses (bachelor’s & master’s) or paid student positions. Note that you can also contact the researchers directly if there is no fitting thesis posted here. For that, take a look at our research areas.

Paid Positions

Privacy-Enhancing Technologies and Differential Privacy: Algorithm implementation and evaluation

(Posted: June 2024 — Flyer)

Do you like the perks of tech but are you tired of big tech companies stealing your personal data? That's where differential privacy comes in. If you want to help us keep your data safe, apply for this HiWi position!

Data analysis has become an important field of study it can, not only improve our daily lives, but can also be used in a variety of ways to analyze data. Communication systems or location-based services are just a few of the many applications that benefit from data analyses today.

While data analysis has many economic and social benefits, most of the data collected contains sensitive information about individuals. To ensure that everyone has the right to privacy, we need to protect the people whose data is being used and make sure it doesn't cause them any harm.

In this HiWi position, you'll work with researchers in the field of database protection, and how to balance useful global statistics about the population with the individual's right to privacy.

Tasks As part of a team of researchers and a current research project, your task will consist of (among other things)

  • Designing, implementing and testing algorithms of current ideas,
  • Evaluating algorithms on real-world data,
  • Writing clear documentation and maintaining usable code bases.

Prerequisites The following are required for this position:

  • Be a KIT student,
  • Primary interest in the topic,
  • Coding skills in Python or similar languages,
  • Basic software development practices (e.g., Git),
  • English communication skills.

If you are interested or have further questions, please contact Àlex Miranda-Pascual.

Theses

A Clustering Trajectory Protection Mechanism with Actual Privacy Guarantees (Bachelor or Master)

(Posted: June 2024 — Flyer)

Did you know that knowing just four location points is enough to identify 95% of the population? While trajectory data is great for analysis, it contains very sensitive information. It can be used to learn where you've been and at what time, when you left home, and when you meet up with friends. So it's really important to make sure that the personal information in a trajectory database is protected.

To keep trajectory data safe, we can use a mechanism that satisfies differential privacy, the go-to privacy notion. For example, we see proposals that claim to introduce a DP clustering mechanism for trajectories. However, none of the DP trajectory clustering mechanisms proposed in the literature actually satisfy DP as they claim.

Thesis Task In this thesis, you'll learn how to identify and fix the problems with existing clustering mechanisms, and you'll also be the first to develop new clustering mechanisms that correctly satisfy DP. You'll also be able to test your mechanisms on real-world data and scenarios.

If you're interested in closing this gap in the field of trajectory protection and keeping your data safe, this thesis is for you.

Prerequisites Primary interest in the topic. Coding skills in Python or similar languages, basic software development practices (e.g., Git), and English communication skills are required.

If you are interested or have further questions, please contact Àlex Miranda-Pascual.

Suppression to Enhance Privacy in Trajectory Data (Bachelor or Master)

(Posted: June 2024 — Flyer)

Did you know that knowing just four location points is enough to identify 95% of the population? While trajectory data is great for analysis, it contains very sensitive information. It can be used to learn where you've been and at what time, when you left home, and when you meet up with friends. So it's really important to make sure that the personal information in a trajectory database is protected.

To keep trajectory data safe, we can use a mechanism that satisfies differential privacy, the go-to privacy notion. For example, we see proposals that claim to introduce a DP clustering mechanism for trajectories. However, none of the DP trajectory clustering mechanisms proposed in the literature actually satisfy DP as they claim.

Thesis Task In this thesis, you'll try to enhance utility by using suppression. You'll adapt suppression to trajectory data to see how they can be used to protect people's privacy when they share their location online. You'll look at how the new mechanisms affect privacy and utility, and describe the settings that provide the most improvement.

If you're interested in developing and deploying new tools to protect your location and keep your data safe, this thesis will be a good fit for you.

Prerequisites Primary interest in the topic. Coding skills in Python or similar languages, basic software development practices (e.g., Git), and English communication skills are required.

If you are interested or have further questions, please contact Àlex Miranda-Pascual.