A Study on ChatGPT Film and Television Subtitle Translation from the Perspective of Three-Dimensional Transformation---Take The Story of Minglan for example

Abstract

Based on the theory of Eco-translatology, this article conducts a study on the subtitle translation of the overseas version of the TV drama The Story of Minglan from the perspective of “three-dimensional transformation,” focusing on the translation generated by ChatGPT, and comparing them with Google Translation, Baidu Translation, and DeepL Translation, aiming to evaluate the effectiveness of ChatGPT in subtitle translation for films and television shows. Firstly, this research employs a quantitative analysis method by using two automatic assessment metrics, BLEU and TER, in order to quantify translation quality. Furthermore, case study and qualitative analysis methods are used to compare and analyze the differences in accuracy, linguistic fluency, and cultural adaptability between the YouTube subtitles of the overseas version of The Story of Minglan and the translation generated by ChatGPT from the perspectives of language, culture, and communication. The results indicate that while ChatGPT translations exhibit some readability, the model still shows limitations in handling ideological issues, cultural transmission, emotional capture, and translation accuracy.