본문 바로가기
  • Home

Differences in Korean-English literary translation among ChatGPT, Papago, and human translators

  • The Journal of Translation Studies
  • Abbr : JTS
  • 2024, 25(2), pp.11-37
  • DOI : 10.15749/jts.2024.25.2.001
  • Publisher : The Korean Association for Translation Studies
  • Research Area : Humanities > Interpretation and Translation Studies
  • Received : April 4, 2024
  • Accepted : June 15, 2024
  • Published : June 30, 2024

Chang-Soo Lee 1

1한국외국어대학교

Accredited

ABSTRACT

The present study explores stylistic and quality differences in English translations of a Korean short story by ChatGPT, Papago, and two human translators. Results of two types of quantitative analysis are reported. A PCA analysis showed that, in the distribution of 80 topmost frequent words, Papago was clearly distinguished from human translators while no such distinction was found vis-a-vis ChatGPT. In a follow-up manually-annotated creativity and quality analysis, Papago was again far from measuring up to human translators with near-zero creativity and numerous errors. ChatGPT also made many errors (about half of Papago’s) but exhibited a measure of creativity of the kinds seen in human translation. ChatGPT’s display of creativity, along with a significant reduction in errors compared to Papago, is attributed to ChatGPT’s ability to consider the context in comprehending the source text and transferring the meaning into the target language.

Citation status

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