Machine translation from Japanese and French to Vietnamese, the difference among language families

Abstract

Although Vietnamese is spoken language of more than 90 million people in the world (in 2014), Vietnamese language is still considered as a low-resourced language. Vietnamese NLP still lacks of resources for text and speech processing, especially research on machine translation for Vietnamese is very rare. This paper presents our first attempt to collect and construct French-Vietnamese and Japanese-Vietnamese statistical machine translation systems. These two different languages, French and Japanese, are less focused in Vietnamese-related machine translation research. The differences between these two languages in comparison with Vietnamese can bring out interesting observations.