본문 바로가기
  • Home

A Search for Quality Improvement of Non-professional Translator Translations through Relay Machine Translation: A Case Study of an Undergraduate Course Practice

  • The Journal of Translation Studies
  • Abbr : JTS
  • 2019, 20(4), pp.83-113
  • DOI : 10.15749/jts.2019.20.4.004
  • Publisher : The Korean Association for Translation Studies
  • Research Area : Humanities > Interpretation and Translation Studies
  • Received : August 30, 2019
  • Accepted : October 8, 2019
  • Published : October 31, 2019

SEUNG HYE MAH 1 Sung, Seung-eun 1

1한국외국어대학교

Accredited

ABSTRACT

The rapid development of machine translation(MT) has recently facilitated the spread of non-professional translations. Non-professionals or amateur translators participate in translation works in a wider range of contexts, emerging as a new distinctive social and cultural phenomenon. Therefore, this research explores ways to help non-professional translators make full use of MT systems. One of the ways suggested in this research is to run MT twice or a relay MT. It is applied to a Korean-English translation practice of 51 undergraduate students majoring in English interpreting and translation. The hypothesis at the practice is that the MT would produce better translation output in English if it goes through Korean-Japanese-English translation process because the quality of MT output depends heavily on accumulated parallel data. It has long been known that Japanese-English translation parallel data have been systematically stored, whereas the history of collecting Korean-English translation data is not as long as that between Japanese and English. The participants compare and analyze two MT outputs, and many of them conclude that the Korean-Japanese-English MT is better to post-edit than the Korean-English MT. They also emphasize that building their translation capacity is as important as knowing how to make use of MT efficiently because ultimately, they are the ones who judge the MT quality and make final decisions.

Citation status

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