Sanskrit and Computational Linguistics
Relive the Webinar Moments
Dr. Pawan Goyal from IIT Kharagpur discusses the development of computational tools for Sanskrit language processing, combining machine learning with traditional linguistic insights to unlock Sanskrit’s vast textual heritage. He outlines efforts to build a neural NLP toolkit—such as SanskritShala—that handles text segmentation, morphological analysis, dependency parsing, and compound detection, showcasing state‑of‑the‑art performance and real-time web-based interaction.
He highlights a significant project on speech recognition in Sanskrit, including a pioneering 78‑hour annotated corpus. The work uses phonetic features and syllable-based units to improve recognition accuracy, overcoming Sanskrit’s linguistic complexities like sandhi and variable spellings arxiv.org . Overall, this video showcases how cutting-edge AI and computational linguistics are making Sanskrit education, research, and digital interaction more accessible and effective.
