This research examines the perceptions of Indonesian migrant workers in Japan regarding two translation techniques, namely Machine Translation (MT) and adaptation translation, specifically in occupational safety vocabulary. This study aims to evaluate the level of clarity and also the effectiveness of communicating occupational safety information through both methods. Research data were analyzed using statistical methods after being obtained from 24 migrant workers using a Likert-scale questionnaire. Although MT is a good starting point, it loses some accuracy if used alone, hence adapted translations are better because they are easier to understand and also culturally relevant. Combining MT with the adaptation process is essential for better workplace safety communication, according to research findings. Research results based on PLS-SEM indicate that adaptation has the most dominant influence on the understanding of safety signs (β = 0.576; t = 2.739; p = 0.006), while MT has no significant effect (β = 0.236; t = 1.148; p = 0.252). The model's predictive capability is classified as moderate (R² = 0.486). These findings indicate that adaptation is more effective than machine translation in helping Indonesian migrant workers in Japan understand occupational safety signs, whereas machine translation alone is not enough. Therefore, combining MT with linguistic and cultural adaptation is important to improve workplace safety communication.
Copyrights © 2026