Speaker: Takako Aikawa, Senior Lecturer in Japanese, Massachusetts Institute of Technology

Takako Aikawa began teaching Japanese at MIT as a Senior Lecturer from September of 2013. She received her PhD in Japanese linguistics from Ohio State University in 1993. Prior to MIT, she worked for Microsoft Research as a computational linguist and a program manager.

Abstract: This lecture addresses challenges that machine translation faces in translating English into Japanese. First, I will provide a brief history of the development of machine translation and explain how a machine can learn mappings between two languages automatically. Second, I will present problems for machine translation in general. Third, I will tackle the question of why the English -> Japanese translation is so hard for machine translation while presenting some linguistic differences between the two languages. Last, I will discuss various types of research that have been done to improve the quality of English->Japanese machine translation.