معرفی
Dr. Daniel Ruffinelli is a Researcher at the Data and Web Science Group (DWS) at the University of Mannheim, affiliated with the School of Business Informatics and Mathematics. His research focuses on Natural Language Processing (NLP), particularly using interpretability methods to study how Large Language Models (LLMs) represent natural language internally. He also works on representation learning and machine learning, with prior PhD research on knowledge graph embeddings under Prof. Rainer Gemulla.
Teaching responsibilities include advanced courses on Text Analytics, Information Retrieval, and Deep Learning, alongside tutorials for Machine Learning, Data Mining, and Large Scale Data Management courses. He has advised numerous MSc and BSc theses since 2018.
Key research highlights include contributions to knowledge graph embeddings (e.g., pre-training strategies and evaluation methodologies) and the development of KGxBoard, an explainable leaderboard system for knowledge graph completion models. Notable recognition includes the Outstanding Paper Award at RepL4NLP@ACL 2019.
His work bridges interpretability techniques with foundational AI research, emphasizing reproducibility through tools like LibKGE, a knowledge graph embedding library.
