Institutional Repository of Vilnius University
Asmeninių duomenų indentifikacija ir anonimizacija naudojant didelės kalbos modelius Personal data identification and anonymization using large language models (LLMs) /
Abstract
dc:descriptionThis thesis explores the use of large language models (LLMs) for anonymizing personal data, addressing the critical balance between data utility and privacy. Focusing on TextWash, ChatGPT, and Bing, it evaluates their effectiveness in identifying and anonymizing personally identifiable information (PII) within texts. Through comparative analysis and the development of an anonymization API, the research showcases the varying capabilities of these LLMs in enhancing data privacy. The findings reveal TextWash's notable precision in data anonymization, assessed using the National NLP Clinical Challenges (n2c2) dataset. Moreover, the newly developed API highlights the practical implications of the study, offering a streamlined, user-friendly solution for real-world data anonymization challenges. This thesis contributes to the discourse on leveraging technology for privacy protection, paving the way for future innovations in the field.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2024
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Sadauskas, Ignas,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD210647911&prefLang=en_US