Aplikasi Pengenalan dan Penulisan Aksara Jawa Berbasis Mobile dengan ResNet-18 Bertema Gastronomi Desa Wisata Gunungsari

Putri, Almira Fidela Soehartanto (2026) Aplikasi Pengenalan dan Penulisan Aksara Jawa Berbasis Mobile dengan ResNet-18 Bertema Gastronomi Desa Wisata Gunungsari. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5025221222-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (10MB) | Request a copy

Abstract

Aksara Jawa sebagai warisan budaya dalam kurikulum muatan lokal Jawa Timur menghadapi penurunan minat belajar akibat metode pembelajaran yang masih konvensional.Penelitian ini mengembangkan aplikasi pengenalan dan penulisan aksara Jawa berbasis mobile Android untuk melatih keterampilan menulis aksara Jawa dasar secara interaktif melalui mekanisme path node tracing dan umpan balik korektif berbasis klasifikasi citra. Sistem menggunakan model klasifikasi citra Convolutional Neural Network berarsitektur ResNet-18 yang dilatih menggunakan dataset aksara Jawa dari Kaggle melalui platform Roboflow. Aplikasi dikembangkan menggunakan Unity dan Figma, serta mengintegrasikan narasi gastronomi Desa Wisata Gunungsari sebagai bentuk pelestarian budaya lokal. Evaluasi dilakukan melalui black box testing, pengujian model klasifikasi, dan System Usability Scale (SUS). Model klasifikasi memperoleh accuracy 91,1%, precision 97,3%, recall 92,2%, dan F1 score 94,2% pada test set sebanyak 156 gambar, dengan confidence threshold optimal sebesar 60%. Pengujian usability terhadap 40 responden menghasilkan rata-rata skor SUS sebesar 80,81 serta respons positif terhadap fitur tracing, feedback, tampilan, story, dan banner Gunungsari. Pengujian edukatif tambahan kepada 13 responden menunjukkan peningkatan rata-rata skor dari 7,08 pada pre-test menjadi 8,77 pada post-test. Hasil penelitian menunjukkan bahwa aplikasi ini berpotensi menjadi media pembelajaran aksara Jawa yang interaktif, fleksibel, dan menyenangkan, sekaligus mendukung revitalisasi budaya digital berbasis kearifan lokal.
=====================================================================================================================================
Javanese script as a cultural heritage in the East Java local content curriculum faces a decline in learning interest due to conventional learning methods. This study developed an Android mobile-based Javanese script recognition and writing application to train basic Javanese script writing skills interactively through a path node tracing mechanism and corrective feedback based on image classification. The system uses a Convolutional Neural Network image classification model with ResNet-18 architecture trained using the Javanese script dataset from Kaggle through the Roboflow platform. The application was developed using Unity and Figma, and integrates the gastronomic narrative of Gunungsari Tourism Village as a form of local cultural preservation. Evaluation was carried out through black box testing, classification model testing, and System Usability Scale (SUS). The classification model achieved 91.1% accuracy, 97.3% precision, 92.2% recall, and 94.2% F1-score on a test set of 156 images, with an optimal confidence threshold of 60%. Usability testing on 40 respondents yielded an average SUS score of 80.81 and positive responses to Gunungsari's tracing, feedback, display, story, and banner features. An additional educational evaluation involving 13 respondents showed an increase in the average score from 7.08 in the pre-test to 8.77 in the post-test. The results indicate that this application has the potential to be an interactive, flexible, and enjoyable learning medium for Javanese script, while also supporting the revitalization of digital culture based on local wisdom.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Aksara Jawa, aplikasi mobile, klasifikasi citra, CNN, ResNet-18, path node tracing, Desa Wisata Gunungsari, Javanese script, mobile application, image classification, CNN, ResNet-18, path node tracing, Gunungsari Tourism Village
Subjects:	L Education > L Education (General)
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55201-(S1) Undergraduate Thesis
Depositing User:	Almira Fidela Soehartanto Putri
Date Deposited:	28 Jul 2026 01:40
Last Modified:	28 Jul 2026 01:40
URI:	http://repository.its.ac.id/id/eprint/138134

Actions (login required)

View Item