Survey on Neural Machine Translation for multilingual translation system

Neural Machine Translation (NMT) has provided promising results in the field of machine translation in recent times. As compared to the previously used rule-based methods or Statistical Machine Translation (SMT), NMT outperforms SMT in quite a few linguist categories. In this paper, we analyze various models, approaches and frameworks used in NMT to find an efficient method to create a translation system and to achieve accurate translation between Indian languages and English.