Perbandingan Kinerja Sistem Neural Machine Translation Opennmt Dan Thumt Dalam Eksperimen Menerjemahkan Bahasa Jawa Ngoko – Krama

Purwaningsih, Leni and , Endang Wahyu Pamungkas, S.Kom., M.Kom, Ph.D (2023) Perbandingan Kinerja Sistem Neural Machine Translation Opennmt Dan Thumt Dalam Eksperimen Menerjemahkan Bahasa Jawa Ngoko – Krama. Skripsi thesis, Universitas Muhammadiyah Surakarta.

BACA FULLTEXT
[img] PDF (Naskah Publikasi)
naskah publikasi - Leni Purwaningsih.pdf
Download (414kB)
[img] PDF (Surat Pernyataan Publikasi)
Surat Pernyataan Publikasi.pdf
Restricted to Repository staff only
Download (228kB)

Abstract

This research was conducted to test the NMT approach in translating Javanese language from ngoko to krama using OpenNMT and THUMT. Both systems were compared based on the accuracy of translation results, which were evaluated using BLEU (Bilingual Evaluation Understudy) scores. The steps of translating Javanese language from ngoko to krama using OpenNMT and THUMT include creating a dataset, building a vocabulary, creating an NMT model, translating test data, evaluating the translation results using BLEU scores, and creating a translation application using the NMT model. The experimental results show that a dataset consisting of a pair of training data containing 1000 sentences and a pair of validation data containing 500 sentences, can produce BLEU scores of 98,55 for OpenNMT and 99,47 for THUMT for the translation of test data that contains sentences found in the training data. Meanwhile, for the translation of test data that contains sentences not found in the training data, the BLEU scores are 2,01 for OpenNMT and 1,6 for THUMT. A Javanese language translation application from ngoko to krama was successfully created using Python programming language and named "Nerjemahake".

Item Type: Thesis (Skripsi)
Uncontrolled Keywords: bahasa Jawa, NMT, penerjemahan, OpenNMT, THUMT
Subjects: T Technology > TZ Technical Information
T Technology > TZ Technical Information > Software. Aplication
Divisions: Fakultas Ilmu Komunikasi dan Informatika > Teknik Informatika
Depositing User: LENI PURWANINGSIH
Date Deposited: 23 May 2023 03:38
Last Modified: 23 May 2023 03:38
URI: http://eprints.ums.ac.id/id/eprint/112415

Actions (login required)

View Item View Item
UMS ETD-db is powered by EPrints 3 which is developed by the School of Electronics and Computer Science at the University of Southampton. More information and software credits.