Multilingual Language Processing in the 21st Century: Lessons Learned and Challenges Ahead
COM1 Level 2
Video Conference Room, COM1-02-13 and via zoom
Zoom link: https://nussg.zoom.us/j/82740525165?pwd=2AKZwxapsbnhbAyrxBmuMbaFxkBaEa.1
Meeting ID: 827 4052 5165
Passcode: 827054
Abstract:
Over the last quarter century, there has been tremendous progress in natural language processing tasks such as machine translation, moving from "eMpTy promises" to an everyday tool. This talk reflects on key drivers for that progress and presents a number of current challenges at the research frontier:
(1) extending machine translation from the text to expressive, prosodically rich speech,
(2) enabling interaction with large language models across all languages equally, and
(3) combining neural and symbolic processes for complex reasoning tasks.
Bio:
Philipp Koehn is a Professor of Computer Science at Johns Hopkins University and formerly held the Chair of Machine Translation in the School of Informatics at the University of Edinburgh. He received his PhD from the University of Southern California, where he worked at the Information Sciences Institute, and completed a postdoctoral fellowship at the Massachusetts Institute of Technology. A leading researcher in machine translation, he published two textbooks in the field and over 200 research papers on natural language and speech processing. He created widely used data resources (e.g., Europarl, Paracrawl) and open source software (e.g., Moses). He founded the Conference on Machine Translation (WMT) and the ACL Special Interest Group on Machine Translation (SIGMT), serving as its president for a decade. His contributions have been recognized with the Award of Honor from the International Association for Machine Translation, and he is a Fellow of the Association for Computational Linguistics. He is general chair of ACL 2026.