Abstract

WHEN SASTRA MEETS ARTIFICIAL INTELLIGENCE: REIMAGINING AYURVEDIC MANUSCRIPTOLOGY

Dr. Akshita*, Dr. Sapna Kaushish

Abstract

Āyurvedic manuscripts represent an extensive repository of knowledge encompassing classical principles, commentarial traditions, therapeutic practices and specialised concepts. However, access to these manuscripts remain limited due to linguistic and textual complexities. Recent advances in Optical Character Recognition (OCR), Handwritten Text Recognition (HTR), Sanskrit Natural Language Processing (NLP) and Artificial Intelligence (AI) offer new possibilities for transforming these manuscripts into searchable, analysable and globally accessible digital resources. The present study explores the potential application of contemporary technologies at various stages of manuscriptology. The study highlights their application as well as limitations including OCR errors, inadequate domain-specific datasets, contextual ambiguity, limitations of general-purpose NLP tools and hallucinations in generative AI. The integration of these technologies and Āyurveda-focused datasets and thorough human validation, can help disseminate traditional wisdom with modern digital science and promote its responsible sharing with a worldwide audience.

Keywords: Manuscriptology, OCR, NLP, Artificial Intelligence, HTR, Digital Humanities, Generative AI.