DharmaBot: An Intelligent Conversational Agent for Hindu Studies Education Based on Large Language Model and Global Hindu Knowledge Base

Abstract

The rapid advancement of artificial intelligence, particularly Large Language Models, has created significant opportunities for transforming academic support systems in religious higher education. Hindu studies education in Indonesia faces persistent challenges in providing accessible and doctrinally accurate consultation beyond classroom hours, particularly in institutions with limited instructional resources such as Institut Agama Hindu Negeri Tampung Penyang Palangka Raya. Students frequently encounter knowledge gaps spanning classical scriptures, philosophical schools, ethical frameworks, and global Hindu practice that instructors cannot address within formal class hours alone. This study presents DharmaBot, an intelligent conversational agent designed to support Hindu Dharma consultation for students in Hindu higher education institutions. Employing a Design Science Research methodology through four stages of problem identification, objective definition, system design, and prototype demonstration, the study produced a system integrating a Large Language Model engine with a Retrieval-Augmented Generation pipeline and a structured Hindu knowledge base covering eight thematic domains including classical philosophy, sacred scriptures, theology, ethics, ritual practice, global Hinduism, and Hindu-science intersections. The system delivers bilingual responses in Indonesian and English with transparent source attribution drawn from verified Hindu scholarly documents. Three demonstration scenarios confirmed the prototype's capacity for accurate, reference-grounded responses to philosophical, theological, and comparative Hindu studies queries. DharmaBot contributes a replicable system design for AI-assisted religious education and offers practical implications for the responsible digital transformation of Hindu higher education institutions in Indonesia and globally.