Examining Machine Translation Systems and Translation Quality using the Back-Translation Method

Abstract

Machine translation systems (MTS) can be used to help bridge the language barrier gap within diverse workplaces. However, someone unfamiliar with the target language may have difficulty knowing if the translation is acceptable to those who prefer the target language. To eliminate this issue, the back-translation method can takesource text, translateit to the target text, and then re-translate it back to the source textto determine the quality of the translation(Miyabe & Yoshino,2015). Tounderstand whether back-translation is a valuable practice, I translated 14 workplace exam questionswere translatedto Spanish and back-translated to Englishusing Google Translate and DeepL.Consulting from three bilingual experts helped determine the quality of the back-translations.Findings show that both MTS generated back-translations that were similar in meaning to the source text and grammatically correct. It is also recommended to use morethan one MTS when determining translation quality.