한국어와 영어의 언어적 차이에 따른 기계번역의 문제점과 그에 대한 포스트 에디팅 방향 제안

This paper suggests how to effectively post-edit the English-Korean machine translation (MT) by examining the impact of the structural differences between the Korean and English languages on the quality of the machine-translated Korean text. First, this study collected English expressions whose equivalence in Korean language are hard to form because of the differences in their respective basic language structure. Based on the Lee’s (2000) model, four topics were chosen: English sentences using inanimate nouns for abstract concepts, English passive constructions using diverse transitive verbs, and English expressions of movement and cause-and-effects. After machine translating the English text into Korean using Google and Papago engines, the study examined whether the machine-translated output meets the lexical, syntactic and semantic requirements of Korean (Lee, 2000; Nunan, 2004). Based on this study, several tips are suggested for fixing problems in the MT of English-Korean text. It is noted that the more different English is from Korea in syntactic and cognitive levels, the less productive MTPE is likely to become. Consequently, the findings stress the importance of MTPE to be a more syntactically competent and cognitively mindful process for the productive MTPE of English-Korean translation.