PhD Project on Pseudonymization and Privacy Protection

Lund University was founded in 1666 and is consistently ranked as one of the world's leading institutions. There are about 47,000 students and more than 8,800 employees in Lund, Helsingborg, and Malmö. We unite in our efforts to understand, explain, and improve our world and the conditions of people.

Lund University welcomes applicants with diverse backgrounds and experiences. Gender equality, equal opportunities, and diversity are fundamental principles for all parts of our operations.

Subject Description

Despite GDPR claims about pseudonymization as an important privacy protection method, there is insufficient evidence from research to trust pseudonymization as a foolproof method for protecting personal identity.

This PhD project will experimentally investigate the effectiveness of pseudonymization techniques to protect the writer's identity across various natural language processing applications, from essays to interactions with robots or AI.

The research challenges focus on the fundamental question: Does pseudonymization work as we expect? This project will systematically evaluate different pseudonymization strategies through rigorous re-identification attacks to establish their actual effectiveness in preserving privacy.

The research is structured around two interconnected themes:

• Pseudonymization Effectiveness Analysis: Systematically evaluate how different identifiers, their number, and combinations affect privacy protection levels in text data. The project will assess which pseudonymization techniques provide adequate protection and under what circumstances they may fail.
• Advanced Re-identification Attack Methods: Develop and utilize traditional attacks (reversing anonymization, information aggregation), newer techniques (graph/node attacks), and new attack methods to comprehensively test the robustness of pseudonymization in both static text (essays, documents) and interactive systems (conversational AI, robotics).

You will employ a methodological approach that includes three types of re-identification studies, each focusing on different categories of pseudonymization, their number, and their combination - in relation to the provided protection levels. You will conduct motivated intruder tests to assess the likelihood of re-identification risks in various contexts, with particular attention to written essays and scenarios for human-robot interaction where personal information may be exposed.

This interdisciplinary approach allows you to evaluate pseudonymization effectiveness across different modalities and interaction types, creating a comprehensive framework for privacy protection in AI systems.

Workplace Description

The position is at Lund University's Department of Computer Science within the research division of Robotics and Semantic Systems (RSS). RSS focuses on research and teaching in AI, machine learning, robotics, and robot learning, human-robot interaction, and natural language processing. Together with the Department of Automatic Control, the division operates RobotLab LTH, which provides access to many types of robots.

You will work in a team with other PhD students and researchers at RSS, within VR Project and WARA ML, forming an interdisciplinary team working at the intersection of machine learning, security, and robotics. The position offers opportunities for international collaboration with leading research institutes and technology companies at the forefront of AI development.

More information about the Department of Computer Science (cs.lth.se (https://cs.lth.se))
More information about RSS (rss.cs.lth.se (https://rss.cs.lth.se))
More information about RobotLab LTH (robotics.lth.se (https://robotics.lth.se))
More information about VR Project (mormor-karl.github.io (https://mormor-karl.github.io))
More information about WARA ML (waraml.org (https://waraml.org))

Job Responsibilities

The main task of a PhD student is to engage in their doctoral education, which includes participation in research projects as well as doctoral courses. The responsibilities also include participation in teaching and other departmental work, but no more than 20% of the working time.

The responsibilities also include:

• Conducting research on privacy-preserving techniques for large language models (LLMs) in robotic applications
• Developing and evaluating methods to detect and mitigate re-identification risks
• Participating in regular research seminars, workshops, and international conferences
• Collaborating with industrial partners in real-world applications and case studies

Eligibility

Basic eligibility for doctoral education is met by those who have

• completed a master's degree or
• completed course requirements of at least 240 higher education credits, of which at least 60 higher education credits at the master's level or
• acquired equivalent knowledge in some other way, within or outside the country.

The specific eligibility requirements for computer science are met by those who have:

• at least 60 higher education credits at the master's level relevant to the research subject, or
• a master's degree in computer science, computer engineering, electrical engineering, information and communication technology, engineering physics, or technical mathematics.

Other requirements:

• Excellent command of English, both spoken and written.
• Experience in machine learning and deep learning, especially transformer architectures
• Knowledge of privacy-preserving techniques (differential privacy, federated learning, etc.)
• Programming experience in Python and relevant ML frameworks (PyTorch, TensorFlow)
• Experience with robotic platforms and frameworks (ROS, etc.)
• Understanding of privacy and security principles in AI systems

For the full advertisement, please see: https://lu.varbi.com/what:job/jobID:813931/ (https://lu.varbi.com/what:job/jobID:813931/)

Welcome with your application!

LTH – Lund Institute of Technology – is the engineering faculty at Lund University. At LTH, we educate people, build knowledge for the future, and work hard to develop society. We create space for brilliant research and inspire the creative development of technology, architecture, and design. Nearly 10,000 students study here. Each year, our researchers – many of whom work in world-leading profile areas – publish around 100 theses and 2,000 scientific findings. A range of research results and student projects are refined into innovations. Together we explore and create – for the benefit of the world.

We kindly decline all contacts from advertisers, recruitment, and staffing companies due to government procurement regulations.