Bridging the world with words: Multilingual and multicultural natural language processing

Despite the rapid advances in Large Language Models (LLM), research efforts have historically focused disproportionately on high-resource languages, particularly English, leaving over 7,000 living languages underserved. We address the fundamental challenge of bridging the gap of low-resource language (LRL) translation in multilingual language models. Low-resource languages are typically characterized by a scarcity of both unlabeled and labeled data, as well as limited tools and models. This talk explores strategies aimed at bridging the gap of low-resource language (LRL) translation in multilingual models, where LRLs are characterised by a limited scarcity of both unlabeled and labelled data, as well as limited tools and models. Speaker(s): Dr. Lee, Room: 313, Bldg: Bergeron Centre for Engineering Excellence, 11 Arboretum Ln, North York, Ontario, Canada, M3N 3A7