Frances Yung

Dr. Frances Yung is a Postdoctoral Researcher at Saarland University's Department of Language Science and Technology within the Department of Computer Science. She has been affiliated with Prof. Vera Demberg's research group since April 2017 and is currently working on the DFG-funded SFB-1102 project "Information Density and Linguistic Encoding," specifically on project B2 "Cognitive modelling of information density for discourse relations." She is pursuing her habilitation, indicating career progression toward a higher academic position in the German university system.

Dr. Yung's research focuses on discourse relations at the intersection of NLP, corpus linguistics, and experimental psycholinguistics. Her work explores how information density affects discourse relation marking through cognitive modeling approaches. She has developed expertise in discourse parsing, resource construction, annotation aggregation, and experimental pragmatics, with particular attention to multilingual aspects of discourse phenomena. Her research combines computational modeling with experimental methods to understand how speakers produce and comprehend discourse relations.

Analysis of Dr. Yung's recent publications reveals a strong focus on discourse relation resources, particularly multilingual corpora like DiscoGeM 2.0 covering English, German, French, and Czech. Her work increasingly incorporates crowdsourcing methodologies and examines how large language models can be leveraged for discourse annotation tasks. She has made significant contributions to understanding the challenges of implicit discourse relation annotation and the biases introduced by different task designs in crowdsourcing environments.

• Active reviewer for major computational linguistics conferences (ACL, EMNLP, NAACL, EACL, COLING, IJCNLP) and workshops since 2016
• Served as area chair for Sigdial 2024
• Regular service on program committees for discourse-related workshops

Dr. Yung has supervised multiple Master's theses on topics related to discourse relations, implicit relation identification, and domain adaptation. Her teaching portfolio includes courses on crowdsourcing linguistic annotations, discourse relations from cognitive and NLP perspectives, and recent advances in discourse processing. She has also served as a teaching assistant for data science and AI courses, demonstrating her commitment to interdisciplinary education at the intersection of computer science and linguistics.