中韓(韓中) 번역을 통해 본 구글번역(GNMT)의 문제

Abstract

In this paper I analyzed the machine translation’s problems through the Chinese-Korean translation, Korean-Chinese translation and Google’s latest translation technology.(GNMT: Google’s Neural Machine Translation System) Machine translation’s problems can be largely categorized as the difficulty of ST’s correction, the difficulty of using translation strategies, the difficulty of TT’s revision and the difficulty of context’s treatment. There is close correlation between the arcane of human language and machine translation’s problems. Surprisingly, human can understand their languages and communicate each other by using their brain in spite of language’s incompleteness. If AI(Artificial intelligence) can compare to human brain, machine can be equal to human in translation. So in this paper, I conclude that ultimately machine can’t reach human’s level in translation. But before long machine translation can produce high quality translation in low context and informative text. This also can make a big impact on language-related fields. So translation and language professionals should prepare for the AI age urgently.