본문 바로가기
  • Home

A comparative study on machine translation and human translation in Korean mimetic words – Focusing on Korean-Chinese translation of modern novels

  • Journal of Chinese Language and Literature
  • 2024, (97), pp.241-274
  • Publisher : Chinese Literary Society Of Yeong Nam
  • Research Area : Humanities > Chinese Language and Literature
  • Received : November 20, 2024
  • Accepted : December 13, 2024
  • Published : December 30, 2024

YoungJi Chung 1

1경북대학교

Accredited

ABSTRACT

This study focused on comparing and analyzing the translation results of Korean mimetic words into Chinese by human translators and three types of machine translation software: Papago, Baidu, and Google. The analyzed work is Kim Young-ha’s “The Empire of Light,” and the human translation is Xue Zhou’s “光之帝國”. There are 153 types of mimetic words in the original text, and the number of times they appear is 396. As a result of analyzing the translation situation of Korean mimetic words, Google Translate omitted the most cases without preserving the meaning of the original text, and also had the most translation errors. In contrast, Baidu omitted fewer words than human translators, and its translation results were also the most similar to human translation. Through this, we can infer that among the three types of machine translators, Google Translate has the least amount of data because deep learning has not progressed as much as the other two translators for Korean mimetic words, while Baidu Translate has the most data because deep learning has progressed the most.

Citation status

* References for papers published after 2023 are currently being built.

This paper was written with support from the National Research Foundation of Korea.