Latin to Sundanese script conversion using Finite State automata algorithm

Abstract

Indonesia is rich in local language. Several of them have their own script, such as Sundanese. Sundanese script is a culture legacy that deserves to be preserved. Sundanese script is syllabic that means each script illustrates the symbol of a syllable, so to change it needs spelling separation. This research proposed a system as solution to learn Sundanese script. The system accepted Latin input which would then be converted into Sundanese script. Latin script was splitted into syllable using the Finite State Automata algorithm with the Non-Deterministic Finite Automata (NDFA) model. Then, transition of syllable separation was represented by transition diagram and guided by the rules of writing and the pattern of Sundanese syllables. We did 6 test cases as a simple black box testing for making sure that the conversion process has done well. Then, we tested the system with 30 respondents using a questionnaire with 5 questions relating to the ease and accuracy of the information provided in the process of converting Latin script to the Sundanese script. With Likert technique for data processing of the questionnaire, obtained the result was around 79.8% of the respondent strongly agree that the information provided was precise, the system was easy to use and easy to understand the Sundanese script.