Dear students, master's thesis opportunities are available on the following topics:
-
Assessing User Privacy Risk in the Unintentional Disclosure of Sensitive Information
This thesis explores methods for evaluating the risk to user privacy resulting from the unintentional release of sensitive information. The work may involve designing frameworks or metrics to quantify this risk, analyzing real-world scenarios (e.g., social media, chatbots, digital assistants), and applying techniques from Natural Language Processing (NLP) to identify and classify sensitive content. The goal is to understand how users unknowingly reveal private data and to develop tools that help mitigate such risks. -
Graph Database Sanitization: Keyword-Based and NLP Model Approaches
This thesis investigates techniques for sanitizing graph-structured databases to ensure sensitive information is protected. Two primary approaches will be considered: keyword-based filtering and more advanced NLP-driven models. The research involves understanding how data propagates in graph structures, defining sanitization strategies that preserve data utility while protecting privacy, and evaluating the effectiveness of these methods in real scenarios.
If you are interested in either topic, feel free to contact me directly.
Applicants should have a solid background and interest in Natural Language Processing (NLP), along with curiosity about security and privacy-related challenges.
Best regards,
Marco Viviani