معرفی
Roxana Pop is a Research Fellow at the University of Oslo, affiliated with the Faculty of Mathematics and Natural Sciences and the Data and Knowledge Management (DKM) research group. Her research focuses on integrating neural methods with symbolic AI approaches to enhance temporal predictions on structured data, particularly through graph neural networks (GNNs).
Her work aims to extract interpretable logical rules from trained neural networks, combining the strengths of data-driven and symbolic AI paradigms. Before her PhD, she conducted research in Natural Language Processing (NLP), specializing in Abstract Meaning Representation (AMR) parsing, a semantic parsing technique.
In 2024, Pop co-authored a publication in the Proceedings of the Seventh Fact Extraction and VERification Workshop (FEVER) titled FactGenius: Combining Zero-Shot Prompting and Fuzzy Relation Mining to Improve Fact Verification with Knowledge Graphs, which explores fact verification techniques using knowledge graphs and zero-shot prompting. Her projects include dynamic link prediction and evaluating the trainability of GNNs for logic-expressible functions.
She is associated with the Data and Knowledge Management group, with research interests spanning Artificial Intelligence, Machine Learning, and Knowledge Graphs. Her profile does not list any scientific awards or advisees.
