An Analysis of the Accuracy of Google Translate in Translating Academic Sentences: A Literature Review

Abstract

The increasing use of machine translation tools in academic contexts has raised concerns about their accuracy in translating complex academic language. This study aims to analyze the accuracy of Google Translate in translating academic sentences through a qualitative literature review. The review synthesized findings from 16 selected studies published between 2014 and 2025 that examined Google Translate or machine translation in academic, educational, and language learning contexts. The studies were selected based on their relevance to translation accuracy, error types, user perceptions, and the use of Google Translate in EFL settings. The data were analyzed using thematic categorization, focusing on grammatical accuracy, lexical and terminological accuracy, contextual and semantic accuracy, translation performance across language pairs, and educational implications. The findings indicate that Google Translate performs relatively well in translating simple and general sentences, especially those with clear structures and common vocabulary. However, its accuracy decreases when translating complex academic sentences, technical terms, culture-bound expressions, and context-dependent meanings. The review also shows that translation accuracy varies across language pairs and is influenced by sentence complexity, linguistic structure, and the availability of language resources. In educational contexts, Google Translate can support students’ comprehension, vocabulary learning, and academic writing, but excessive dependence on the tool may reduce students’ critical language awareness. Therefore, Google Translate should be used as a supportive tool accompanied by critical evaluation and human post-editing. Future studies are recommended to test Google Translate using real academic sentences written by Indonesian EFL students.