Shastra as a Rulebase

Potential Research Projects at IITs

Dr. Sai Rama Krishna Susarla

Introduction

A number of Bharatiya shastras provide guidance on handling certain scenarios based on Vedic model of reality. Examples include Ayurveda, Jyotisha, Vastu, and Dharma Shastra. This guidance is expressed in the form of conditions and actions. These condition-action pairs are called rules. For practitioners, it is very convenient if shastra granthas are presented explicitly in the form of rules, as it helps them directly in their practice. For example, the symptoms of a disease for which a medicine applies is expressed as a rule. A dharma shastra vidhi is prefixed by the scenario to which it applies. However, shastras express such rules in the form of shlokas logically connected with higher-level semantic relations called Tantrayuktis. To correctly interpret a shastra, one needs to not only have knowledge of the language, but also vocabulary, technical terminology (Shastra Paribhasha), and semantic conventions such as tantrayuktis. Above all, one needs an intuitive ability to choose an appropriate rule among possibly hundreds of matching rules for a situation. The main difference between an expert and a novice in a shastra lies in their knowledge level in the above aspects. If the expert interpretation of a shastra is available for novices to consult, it will greatly enhance the effectiveness of shastra practice in the society. Computing and AI can be used to make this possible.

Potential Ph.D. Projects

We propose research projects at premier technical institutes with computing and domain expertise to transform Bharatiya shastras in various disciplines to be converted into automatically processable rulebases for deploying Bharatiya knowledge effectively for the benefit of modern society. As a next step, we can enhance such rulebases with guidance of expert practitioners to help train beginners. We propose piloting this effort with Brihat Trayii of Ayurveda. This requires –

1. Transforming Ashtanga Hridayam, Charaka Samhita and Sushruta Samhita into grammatically tagged text
2. Concept mapping the text at the Vakya level and concept (Vishaya) level
3. Identifying condition-action pairs from the concept map
4. Developing a rule-based interpreter and a physician assistant application using it.

This project requires collaboration between the following parties:

1. Sanskrit and Shastra scholars (Samskrit universities)
2. Ayurveda experts
3. Computer science and AI experts (IITs)
4. Digitization platform providers (SKF)