Murathan Kurfalı

Murathan Kurfalı is an active researcher at Uppsala University's Department of Linguistics and Philology within the Faculty of Languages, specializing in Natural Language Processing with a strong focus on Grammatical Error Correction (GEC), discourse analysis, and multilingual applications. His work bridges computational linguistics with practical language education tools, particularly for Swedish language processing.

His research interests center on developing and evaluating NLP systems for language learning, with significant contributions to GEC evaluation methodologies using Large Language Models. He has pioneered reference-free evaluation metrics through LLM-based post-editing and created benchmark datasets like SweSAT-1.0 for assessing Swedish language models. His work extends to climate change applications through ClimateEval, a comprehensive NLP benchmark for climate-related tasks.

Analysis of his publication trends shows consistent advancement in discourse analysis (particularly implicit relation classification and connective detection) alongside growing emphasis on climate informatics and Swedish language model evaluation. His methodological approach favors resource-efficient techniques like gradient boosting for connective detection while leveraging LLM capabilities for complex tasks.

As a collaborative researcher, he frequently partners with Uppsala colleagues including Joakim Nivre and Mats Wirén, contributing to major workshops like BEA, ClimateNLP, and NLP4CALL. His work demonstrates strong translational impact from theoretical NLP to practical language education and climate science applications.