A Comparison of Google Translate and Human Translation Results in Students Thesis Abstracts: Accuracy and Readability

Abstract

Amidst the growing use of machine translation in academic settings, this study investigates the comparative quality of thesis abstract translations produced by Google Translate and professional human translators. Framed within the context of Translation Studies and linguistic evaluation, it aims to examine how each approach performs in terms of accuracy, fluency, and readability. This research employed descriptive qualitative methods to compare the translation results between Google Translate and human translation. By focusing on lexical precision, syntactic coherence, and contextual appropriateness, the study reveals that while Google Translate, powered by Neural Machine Translation (NMT), exhibits significant improvement in structural accuracy and terminological consistency, it frequently struggles to handle idiomaticity, pragmatic shifts, and nuanced academic discourse. Human translations, by contrast, consistently demonstrate superior contextual sensitivity, naturalness, and discursive flow. These findings highlight the current limitations of machine translation tools, such as Google Translate, in capturing the complexities of academic language and reaffirm the continued importance of human mediation in achieving high-quality scholarly communication.