Abstract
Balinese Script and other inheritance local languages are facing extinction signalment since their use has already been replaced by their national languages, which are simpler and more practical. Bali province has already had governor regulations to preserve this local wisdom through many efforts in Bali, including conducting compulsory local content subject Balinese language (covering Balinese Script) from primary to high school, requiring institution's nameplates to be written in Balinese Script along with different languages, etc. This collaborative study, a joint effort between language and computer science, has exposed the technological side of support for preservation efforts. Through the mobile application from a smartphone, this study revealed the advantage of using this proposed Information Technology application for 21st-century learning of Balinese Script, including paperless material to conduct green-and-sustainable-oriented learning and efficiency to grab knowledge in real-time related to the transliteration and translation aspects from Latin text input. That effort was considered the main contribution to this research area.
Introduction
The Balinese script, a complex Abugida system, is currently facing an "extinction signal." As global and national languages dominate daily communication, traditional scripts are often sidelined due to perceived complexity. While the Bali provincial government has mandated the script's use in education and public signage, there is a critical need for modern technological tools to make the script accessible and relevant to 21st-century learners.
This research presents a mobile-first framework designed to bridge the gap between ancient heritage and digital literacy. By developing the ToAksara (Latin to Balinese) and ToLatin (Balinese to Latin) applications, the study provides a "green" and efficient solution for script preservation.
Methodology: Efficiency Through Finite-State Machines
Transliterating between the Latin alphabet and Balinese script is challenging due to linguistic rules like pasang pageh (exceptions) and scriptio continua (lack of spaces). To solve this, the researchers implemented a Finite-State Machine (FSM) algorithm.
Unlike heavy AI models, FSMs are computationally lightweight, making them ideal for mobile devices with limited resources. The FSM manages state transitions to ensure that character combinations are converted into proper Balinese glyphs, handling complex grammar such as nasal prefixes and "sound killers" (pangangge tengenan). The applications were developed using the Flutter framework to ensure cross-platform compatibility and utilize Unicode-compliant fonts like Noto Serif Balinese for accurate rendering.
Datasets and Knowledge Integration
The primary knowledge base for the transliteration engine is a digitized version of the Balinese-Indonesian Dictionary (Anom et al., 2013). This source served as the ground truth for validating the FSM model's linguistic accuracy.
In addition to the original dictionary data, the researchers added English translations. This expansion makes the tool useful not only for local students but also for international researchers and tourists, further promoting the global visibility of Balinese culture.
Key Results and Impact
The applications were successfully tested on Android 13 devices, demonstrating high performance on standard hardware. The system achieved robust accuracy across 16 distinct linguistic categories, including basic syllables and complex abbreviations.
Beyond technical performance, the project promotes sustainable education. By moving from paper-based dictionaries to mobile apps, the study supports "green" learning initiatives. The tools have already been introduced to high school teachers in Bali, such as at SMKN 2 Tejakula, where they were recognized for their potential to modernize classroom instruction.
Conclusion and Future Work
This study demonstrates that traditional computer science models like FSMs remain highly effective for specialized linguistic tasks in the mobile era. Future developments will focus on personalizing the learning experience with user accounts and incorporating gamification to better engage younger generations. By integrating cultural wisdom with mobile technology, this framework offers a scalable model for preserving other endangered regional scripts.
Summary updated