Comparative Analysis of Translation Errors in F The Prom Subtitles by YouTube and DeepL

Abstract

Accurate subtitle translation is essential in films, as it influences viewers' comprehension and connection with cultural nuances. This study addresses the challenges in automated subtitle translation, focusing on translation errors in idiomatic expressions, abbreviations, slang, onomatopoeia, and address terms.  This research addresses a gap in the field by directly comparing the two automated translation tools using Koponen's (2010) error classification theory,  highlighting their limitations in handling idiomatic expressions and cultural nuances, thereby providing insights necessary for enhancing automated translation algorithms. The methodology applied is descriptive qualitative. This study evaluates and compares translation errors in subtitles generated by YouTube Automated Translation and DeepL for F The Prom movie, highlighting that both systems have significant limitations in accurately translating idiomatic expressions, slang, and cultural nuances. YouTube often misses culturally specific terms, while DeepL struggles with idioms and slang, indicating a critical need to enhance translation algorithms to improve accuracy and the viewing experience, particularly for non-native speakers.