MESIN PENERJEMAH BERBASIS NEURAL: STUDI KOMPARATIF TENTANG, KOMPETENSI DAN PERSEPSI MAHASISWA SERTA KUALITAS DALAM PEMBELAJARAN TRANSLASI = NEURAL MACHINE TRANSLATION: COMPARATIVE STUDY ON THE STUDENTS’ COMPETENCE, PERCEPTION AND QUALITY IN TRANSLATION CLASSES


MAHARIDA, . (2026) MESIN PENERJEMAH BERBASIS NEURAL: STUDI KOMPARATIF TENTANG, KOMPETENSI DAN PERSEPSI MAHASISWA SERTA KUALITAS DALAM PEMBELAJARAN TRANSLASI = NEURAL MACHINE TRANSLATION: COMPARATIVE STUDY ON THE STUDENTS’ COMPETENCE, PERCEPTION AND QUALITY IN TRANSLATION CLASSES. Disertasi thesis, Universitas Hasanuddin.

[thumbnail of F013211006-3gsY9qxtjiKv6bHp-20261014091923.pdf] Text
F013211006-3gsY9qxtjiKv6bHp-20261014091923.pdf
Restricted to Repository staff only

Download (4MB)

Abstract (Abstrak)

This study addresses that gap by exploring students’ competence and perceptions in using NMT, while also evaluating the accuracy of their translations. The goal is to understand how the characteristics of these tools influence learning outcomes and the precision of translated texts. Adopting a mixed-methods convergent design, the research combined quantitative data from Likert-scale questionnaires with qualitative insights from interviews, alongside an analysis of students’ translation outputs. Thirty English Department students from two universities in Makassar were purposively selected as participants. Findings show that GT remains the preferred choice for quick lexical lookups, while ChatGPT is favored for longer, paragraph-level translations requiring coherent phrasing and contextual nuance. Three user profiles emerged—critical, strategic, and passive—with critical users producing the most polished, high-quality translations. Accuracy tests revealed that ChatGPT generally outperformed GT, though both tools struggled with idiomatic expressions, cultural references, and specialized terminology. The study concludes that NMT tools, when used critically and strategically, can enhance vocabulary, comprehension, and translation quality. However, their limitations reinforce the ongoing need for human judgment and post-editing in academic translation. Thoughtfully integrating these tools into teaching can build students’ digital literacy and linguistic awareness, positioning them as valuable aids rather than replacements for human expertise.

Item Type: Thesis (Disertasi)
Uncontrolled Keywords: Neural Machine Translation, Google Translate, ChatGPT, Translation Accuracy, Digital Literacy
Subjects: P Language and Literature > P Philology. Linguistics
Divisions (Program Studi): Fakultas Ilmu Budaya > Ilmu Linguistik
Depositing User: Unnamed user with username pklpkl
Date Deposited: 10 Nov 2026 05:02
Last Modified: 10 Nov 2026 05:02
URI: http://repository.unhas.ac.id:443/id/eprint/61751

Actions (login required)

View Item
View Item