Analysis of Machine Translation Performance on Translating Informative Text from English into Indonesian

Abstract

This study aims to investigate the performance of 6 machine translation. The text translated was informative text from English into Indonesian. The document taken from 48 students paper in semester final test. The research design is descriptive qualitative and content analysis approach. The data obtained from the students translation result in final test, observation, and interview. In analyzing the result of translating, there were three categories: grammatical structure, cultural words, and mechanic writing (composition writing). The result shows the performance of 6 machine translations: Google translate (GT), DeepL, Yandex, Systran, Udictionary, Microsoft translator, and itranslate on grammatical structure analysis were understanable related to meaning because the language is news report in formal language and reporting facts. However, some roles of language were changes such as: tenses, word formation, active/passive, singular plural, article, and auxiliary verbs. There was no example in cultural words and mechanic writing because the form of language is news report. The result of observation indicated that the students already apply technology in translation by utilizing MT in translating text. Furtthermore, in the result of interview implied that the usage of MT can give an assistance in translating text from English into Indonesian especially for informative text.