A Survey of Low Resource Neural Machine Translation

Machine translation is one of the important research directions in the field of artificial intelligence and natural language. It has important scientific research value and practical value. Neural machine translation (NMT) is a new machine translation model which uses neural network to realize the transformation from source language to target language. In recent years, neural machine translation method based on sequence to sequence has major breakthrough in machine translation task, in most of the language of the translation effect beyond the effects of statistical machine translation, but for the translation under the condition of scarce resources, neural machine translation has yet to achieve the desired effect, this paper mainly introduces several effective methods and models under low resource conditions, as well as their problems and challenges, and the existing problems and challenges. Then look forward to the future research direction and development trend.