SEMANTIC ERRORS FOUND IN THE TRANSLATION OF TASYA FARASYA’S TIKTOK ACCOUNT

This study investigates the quality of machine translation on TikTok, focusing on the accuracy and effectiveness of auto-generated captions in usergenerated content by examining the semantic errors in the translation product. The primary objective was to identify the types of semantic errors produced by TikTok Machine Translation, particularly in Tasya Farasya's TikTok account and its relation to the translation procedure conveyed in the translation process. Employing a descriptive qualitative approach, the research analyzed translation errors produced by TikTok's machine translation, categorizing them according to Sayogie's (2014) theory of semantic aspects that is supported by the theory proposed by Kridalaksana (2008), including grammatical meaning, contextual meaning, and referential meaning and classifying the translation procedures using the theory by Vinay and Darbelnet (2000). Through an analysis of semantic errors and translation procedures in a sample of Tasya Farasya's TikTok videos, the study reveals significant challenges in accurately conveying contextual meaning. The predominant use of literal translation leads to frequent semantic errors, particularly in contextual meaning, indicating a need for further development and refinement of machine translation algorithms.