This paper studies the translation quality of phrasal verbs generated by Auris AI in the Cartoon Scary Girl (2023) from the terms of accuracy, acceptability and readability. The aforementioned techniques are then applied to translation and the quality of a translated text measured using Nababan's framework. The researcher listed phrasal verbs systematically into a table, which was analysed according to the guidelines given in Spradley theory. As a result, despite Auris AI's claims of achieving up to 99% accuracy with high-quality transcriptions, the study found that the overall translation quality was medium with the average score is 2,24 out of 3,00 or 74,6%. It shows that Auris AI's high claim for translation accuracy is really just a kind of image appealing to the customers. The assessment incorporated three terms: accuracy, which scored 2.25 or 75%; acceptability, which also scored 2.25 or 75%; and readability, which scored 2.24 or 74.6%. These results showed that the translations often do not convey the intended meaning, lack naturalness, and have only a moderate level of clarity and comprehensibility.
Copyrights © 2024