This study examines the translation of Culturemes in Chinese–Korean artificial intelligence (AI) literary translation and proposes practical post-editing (PE) guidelines to improve translation quality. Using a Chinese novel as the source text, the s...
This study examines the translation of Culturemes in Chinese–Korean artificial intelligence (AI) literary translation and proposes practical post-editing (PE) guidelines to improve translation quality. Using a Chinese novel as the source text, the study analyzes how Culturemes are rendered in ChatGPT’s Korean translations and identifies cases where post-editing is required due to insufficient equivalence at the lexical, sentential, or contextual level.
Based on an integrated classification framework, a total of 352 Culturemes were identified in the source text, of which 181 items (51.42%) were classified as requiring post-editing. Quantitative analysis shows that social-specific terms exhibited the highest need for post-editing, while idiomatic expressions demonstrated relatively stable translation performance. Drawing on Aixelá’s preservation–substitution framework, this study applies eight post-editing strategies, including transliteration, intratextual and extratextual explanation, paraphrasing, domestication, generalization, and omission, to representative translation examples.
To evaluate the effectiveness of the proposed strategies, a survey was conducted with 30 native Korean speakers. The results indicate a consistent preference for post-edited translations over raw AI outputs, suggesting that the proposed strategies contribute positively to translation acceptability. Based on these findings, the study presents a set of practical post-editing guidelines for Cultureme translation, offering empirical support for improving the quality of Chinese–Korean AI literary translation.