REGIONAL WRITING SYSTEM PARITTA AND DHAMMAPADA APP WITH CHARACTER CLASSIFICATION USING CONVOLUTIONAL NEURAL NETWORK ON TENSOR FLOW LITE

Abstract

Globalization is the mainstream of this era in 21st century characterized by rapid development of information technology. People now can access readily available information on internet much faster than before with smartphone which can be brought anywhere and easy to use. Not just for accessing information, smartphone can be used for various activities, such as playing games, watching movies, listening music, online shopping, studying, or even for religious and customs purposes all thanks to the wide range of application made by various developer. Owing to such positive effect, globalization also has made people integrated more rapidly and globally, hence more and more people especially in Indonesia are disregarding and even forgetting its own traditional culture. This paper will discuss the development and usage of an application as form of an e-book to help Buddhist to pay respect and take refugee to Buddha by chanting Dhamma also to learn the meaning of it in an appropriate place and time also to make Buddhist in Indonesia remember their culture through the usage of regional writing system. The application will be made multi-platforms with Android as its main supported platform with an innovative feature such as control panel for different customization of fonts, chanting sound of verse, favourite, search features, and intelligent Javanese character classification recognition feature using convolutional neural network with a simple, but attractive and modern user interface design.