Lunch at 12:30pm, (in-person) talk at 1pm, in 148 Fitzpatrick

Title: Multilingual Machine Translation in Language Diversity

Abstract: Multilingual machine translation enables one system to do translation from multiple source languages to target languages, greatly reducing deployment costs compared with conventional bilingual systems. To accommodate the language diversity, the multilingual systems have been actively employed in a translator service such as Microsoft Translator (https://translator.microsoft.com/). The multilingual research, however, is often limited to English-centric translations and a few works has been studied for direct translation from a non-English language X to another non-English language Y. In this talk, I will introduce our recent research projects on how to build a multilingual translation system that serves arbitrary X-Y directions and also discuss challenges in deploying such systems.

Bio: Akiko Eriguchi, PhD. Akiko is a Senior Researcher at Microsoft. Her research interests lie in multilingual NLP and deep learning. With her position in the Microsoft Translator team, she has developed MT systems and multilingual NLP applications. She is also a 2021-2022 co-organizer of the Workshop on Asian Translation. As a part of D&I volunteering work, she serves on the non-profit board of Women in Machine Learning, Inc., to increase the impact of women in machine learning in the community. Prior to joining Microsoft, she was a Research Fellow (DC 1) at the Japan Society for the Promotion of Science. She obtained her doctoral degree from the University of Tokyo, Japan.