The Application of ChatGPT in Literary Translation: A Case Study from Thai to Chinese

Natural Language Processing (NLP) is rapidly evolving, yet it still faces challenges in comprehending deep meanings and emotions, handling diverse languages and subtle linguistic nuances. Literary text translation remains particularly tough for machine translation, with limited research focusing on Thai-Chinese translation. This study investigates the Thai biographical essay '(sic),' (Mother Nao Hua) comparing quantitative and qualitative aspects of translations by Google Translate, Baidu Translate, and ChatGPT. The goal is to explore the difficulties faced in Thai-Chinese machine translation and how to issue clearer commands for literary translations. Quantitative analysis shows ChatGPT's translation as the best, followed by Google and Baidu Translate. Modified instructions improved ChatGPT's translation quality. Qualitative analysis reveals reasons for translation quality variance and machine translation's literary text issues. It finds challenges in Thai to Chinese translation, suggesting ChatGPT instruction refinement can improve quality. The study proposes three future research directions.