Academic Text Quality Improvement By English Department Students of University of Mataram : A Study on Pre-Editing of Google Neural Machine Translation

Abstract

This study aims to determine the quality of students in pre-editing academic text input to GNMT. The participants in this study were students of the English Language Education Study Program, Faculty of Teacher Training and Education, the University of Mataram, who took the "Translation and Interpreting" course in semester five of the 2021/2022 Academic Year. This data was collected from assignments completed by 20 students. Supporting information is collected through observation by sitting while learning takes place. Data were analyzed using content analysis procedures such as identifying, categorizing, describing, and explaining. The results of this study indicate that almost all students have good quality results in pre-editing, but it could be better, and some students fail to pre-edit text. The pre-edited output looks like a revised version of the text. Pre-editing shows how the source text changes, especially in language structure, word choice, and punctuation. The good or bad quality of the GNMT translated text represents the student's ability in pre-editing the source text. Thus, the more effort is put into pre-editing the text context, the more likely it is to produce text with better translation quality by GNMT.