A Comparative Study of Religious Scriptures Using Natural Language Processing

Abstract

Over the past few years, Natural Language Processing (“NLP”) has emerged as a powerful tool and has enabled computational analysis of texts by offering insights into the subtleties of language, emotion, and thematic frameworks. This research paper employs NLP strategies such as topic modelling and sentiment analysis to compare translations of three religious scriptures: the Bhagavad Gita representing Hinduism, Quran representing Islam and the Bible representing Christianity. Before carrying out the tests, text was pre-processed and cleaned to ensure that the most optimum results were obtained. Topic modelling uses algorithms such as Latent Dirichlet Allocation to find prominent themes while sentiment analysis makes use of an NLTK VADER sentiment module ‘Sentiment Analyzer’ to interpret emotional undertones in text. This research paper finds that the texts share similar views on topics such as generosity, devotion to God among others and have differing opinions on themes including sacrifice and violence. It is also interesting to note that while the religions of the Bhagavad Gita and Quran (Hinduism and Islam respectively) have been pitted against each other for centuries in countries such as India, they share several similar principles and ideologies.