معرفی
Benjamin Roth is a Professor at Saarland University holding dual affiliations in the Faculty of Computer Science (Research Group Data Mining and Machine Learning) and the Faculty of Philological and Cultural Studies (Department of European and Comparative Literature and Language Studies). His research focuses on natural language processing, large language models, machine learning, and computational linguistics. He leads multiple active research projects including 'Understanding Language in Context' (2025–2033) and 'Linguistic Methods for the Detection of Implicit Abuse' (2024–2027). Notable collaborations span cross-functional studies on LLM behavior, clinical text analysis, and knowledge graph integration. He actively participates in conference organization and has presented at venues like the Konferenz zur Verarbeitung natürlicher Sprache (2024).
His work emphasizes methodological innovation in weak supervision, model calibration, and multimodal reasoning. Recent contributions include studies on persona effects in LLMs, specification overfitting mitigation, and zero-shot temporal relation extraction. He has co-authored over 20 peer-reviewed publications since 2020, with a focus on advancing ethical AI, model interpretability, and NLP education.
