Shin & Chon (2022). 기계번역의 오류 양상과 사후수정(post-editing)이 L2 쓰기 적절성에 미치는 영향

[Dongkwang Shin]

신동광, 전유아. (2022). 기계번역의 오류 양상과 사후수정(post-editing)이 L2 쓰기 적절성에 미치는 영향. 영어교과교육, 21(1), 121-140.

Shin, Dongkwang & Chon, Yuah Vicky (2022). Machine translation errors and the effect of post-editing on the appropriateness of revised L2 texts. Journal of the Korea English Education Society, 21(1), -140.

Although machine translation (MT) such as Google Translate (GT) may be used by second language (L2) learners to solve lexical and syntactic problems for L2 writing, the use of MT requires L2 learners to notice errors in MT output and post-edit MT errors. The purposes of the study were to 1) analyze the types of MT errors; 2) examine how post-editing of MT errors would have an effect on sentence adequacy scores of post-edited texts; and 3) examine any differences between the MT output and post-edited texts on measures of lexical and syntactic complexity. To this end, 57 L2 university learners were asked to compose a text in L1, translate it with GT, and post-edit the text for any mistranslations, ungrammatical sentences, or incomplete expressions. The MT output produced a total of 931 sentences of which 600 sentences (64.5%) were without any errors. Three hundred and thirty-one sentences had errors in the areas of mistranslation (65.56%), ungrammaticality (16.11%), missing word (16.94%), and extra words (1.39%). When the effectiveness of post-editing was examined, the total mean of sentence adequacy scores was found to have improved significantly in comparison to those of the machine translator. Examination of lexical and syntactic complexity measures indicated an increased number of sentences but a drop in the number of coordinate phrases in the post-edited texts. The results are discussed as to how MT can be used to help learners improve their L2 writing skills.

[machine translation/error types of MT/post-editing/Google Translate/기계번역/기계번역 오류 유형/사후수정/구글 번역]

6. 신동광 전유아.pdf

712.36KB