ScholarWorks@숙명여자대학교

초록

This study aims to examine the translation strategies for effectively translating headlines when post-editing results from machine translation engines and generative AI focusing on the unique characteristics of game texts based on interactivity. For this study, translations of game content from Need for Speed: Unbound and Real Racing 3 were compared across Human Translation (HT), DeepL, Google Translate, Papago, and ChatGPT. The results revealed that human translators employed literal translation, transliteration, and context-based translation, while machine translations predominantly used literal and transliteration methods. Among the platforms, Google Translate used literal translation the most, followed by Papago and DeepL. Conversely, transliteration was most frequently used by DeepL, with Google Translate using it the least. Consequently, while human translation accounted for approximately 25%, the absence of such contextual consideration in machine translations suggests that they fail to capture the contextual nuances of game texts.

키워드