상호문화적 관점을 반영한 AI 번역의 프롬프트 구성 방안- 한국어 ‘환영’과 중국어 ‘欢迎’을 중심으로-

초록

This study aims to explore ways to utilize generative AI translation prompts from an intercultural perspective. Generative AI can function as a tool for inquiry through meaning negotiation and as a collaborative partner in problem-solving. Accordingly, this study proposes a four-step procedure for constructing translation prompts that contribute to understanding linguistic differences arising from intercultural perspectives. The proposed four steps are as follows: (1) identify problematic situations caused by cultural differences, (2) analyze and compare how these differences manifest in various aspects, (3) investigate the underlying causes of the differences, and (4) construct prompts that can yield translation results reflecting intercultural elements. To verify the effectiveness of the prompts, this study compared the translation outputs of AI-based translators (Papago and Google Translate) and a generative AI model (ChatGPT) using the Korean term “환영” and the Chinese term “欢迎” as examples. The analysis revealed that AI-based translators showed limitations in adequately reflecting contextual and discourse-functional elements beyond the basic meaning. In contrast, by applying the step-by-step prompts proposed in this study, ChatGPT was able to produce more meaningful translation results by mediating linguistic and sociocultural differences. These findings confirm that the AI translation prompts proposed in this study can effectively contribute to translation practice from an intercultural perspective and to the enhancement of intercultural understanding.

키워드