An AI-Integrated Intelligent Health Advisory System with Machine-Learning-Based Ayurvedic Formulation Recommendation

Information Science Engineering

AyurAid is a web-based intelligent health advisory system designed to integrate classical Ayurvedic knowledge with modern artificial intelligence methodologies. The system addresses the challenge of limited accessibility to reliable Ayurvedic guidance by enabling users to input symptoms in natural language and receive personalized formulation recommendations along with contextual lifestyle advice. AyurAid incorporates multiple computational layers, including symptom normalization, fuzzy similarity matching, TF-IDF-based document vectorization, and cosine-similarity-driven ranking to identify the most relevant Ayurvedic formulations. In addition, the platform employs a domain-constrained AI advisory model to generate safe, user-friendly explanations grounded in retrieved formulations. The backend architecture is built using a Flask API that manages authentication, symptom preprocessing, recommendation generation, and advisory interactions, while a lightweight web frontend ensures smooth user experience. The recommendation engine was evaluated using curated datasets containing more than 1200 Ayurvedic formulations and nearly 800 symptom descriptors. Experimental results demonstrate strong performance, with Precision@5 of 0.82, Recall@5 of 0.76, and Mean Reciprocal Rank of 0.71, reflecting both accuracy and ranking effectiveness. User studies further indicate high satisfaction with the clarity and usefulness of AI-generated advice. AyurAid’s design emphasizes safety, transparency, and reproducibility by integrating retrieval-based grounding, conservative advisory constraints, and structured evaluation metrics. While the current version focuses on single-session interactions, the system’s modular architecture enables future integration of patient history, hybrid recommendation approaches, and domain-fine-tuned language models. Overall, AyurAid showcases a promising fusion of traditional healthcare principles and contemporary AI techniques, offering a scalable pathway for enhancing accessibility to personalized Ayurvedic health support.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.