Exploration of the ideal sanskrit grammar for natural language processing: An analysis of laukik and Vedic Sanskrit

Abstract

Prior to the writing of this paper, there has been a lot of research into the applicability of Sanskrit as a language for programming. This further expanded into research about how applicable it is for Natural Language Processing (NLP). Reasons cited include its syntax and almost mathematical grammar, along with its lack of exceptions. This made Sanskrit a major candidate in the field of natural language processing. However, since then, research in the field has not progressed further, and we are yet to see Sanskrit being used in natural language processing. Furthermore, Sanskrit is a language with a rich history, dating back several thousand years, and it has evolved a lot since its conception. This gave way to even more possibilities of the usage of Sanskrit, since there were ancient and modern forms of the language. Therefore, researching the ideal version of Sanskrit for usage in Natural Language processing was chosen as a topic, and Vedic(ancient) and Laukik(modern) versions of the language were compared. After conducting research, it was found that Laukik Sanskrit seems to be a more refined version of the language. It has eliminated several ambiguities, and streamlined the grammar. This led to the belief that Laukik Sanskrit is more suitable for Natural Language Processing.