Comparing the quality of Vietnamese-Korean translations between AI and human translators in the genre of journalistic texts

Abstract

Abstract: The study aims to compare the translation quality of Vietnamese-to-Korean journalistic texts
produced by artificial intelligence (AI) and human translators. The question is whether AI has achieved a
quality level equivalent to or superior to human translators in this specific text genre? The study employs a
combination of qualitative and quantitative methods: six Vietnamese news articles were translated by AI
and by five professional translators working in Korean organizations, and subsequently evaluated and
compared by three native Korean lecturers. The evaluation criteria included (1) Accuracy, (2) Naturalness
and Fluency, (3) Journalistic style, (4) The ability to convey cultural implications. The research findings
indicate that AI excels in accuracy, stability, maintaining consistency of terminology, and conveying
complete information. While human translators may make errors such as incorrect or missing translations,
they have an advantage in using flexible and creative language, especially in article titles and emotionally
expressive phrases. This difference stems from the technical characteristics of AI and the difficulties human
translators face when translating into a target language which is their foreign language.
Key words: AI translation; human translator; Vietnamese - Korean Journalism; Translation quality;
Post-editing.