Entity Networks in Vaiṣṇava Purāṇas: A Computational Study of the Viṣṇu- and Narasiṃha-Purāṇa

This study applies computational text analysis to two Purāṇas—the Viṣṇu-Purāṇa and the Narasiṃha-Purāṇa—demonstrating how digital methods can illuminate theological and narrative structures within Sanskrit texts. Using corpus tokenisation, entity recognition, and co-occurrence networks, the project quantifies the distribution of deities, persons, and places across the texts. The results reveal contrasting emphases: the Viṣṇu-Purāṇa invokes Viṣṇu predominantly through epithets within dialogic frames, whereas the Narasiṃha-Purāṇa favours explicit naming of Viṣṇu alongside a dominant oral-recitation voice (sūta). Place references are sparse in both, underscoring their mythic rather than geographical orientation. Beyond these specific findings, the study contributes a replicable methodology that bridges Data Science and Indian Knowledge Systems (IKS), offering a foundation for future digital-humanities engagement with the Purāṇic corpus. The work highlights how computational tools, when sensitively applied to traditional Sanskrit materials, can open new avenues of interpretation and scholarly discussion.