International Journal of Scientific Research and Engineering Development

Traditional online symptom checkers provide static results based on predefined rules and lack emotional awareness or contextual understanding of users’ health descriptions. This research introduces an AI-Based Health Checker System (AI-HCS) that integrates Natural Language Processing (NLP), Machine Learning (ML), and Medical Knowledge Graphs to analyze user symptoms and suggest probable health conditions, remedies, and nearby doctors. The system accepts user input in natural language (typed or spoken) and uses a transformer-based model (e.g., Gemini or GPT API) to identify likely illnesses. The architecture includes four modules — Symptom Analyzer, Disease Predictor, Remedy & Doctor Recommender, and Feedback & Learning Engine. Testing on 500 symptom queries achieved an average accuracy of 89.4% and reduced user misclassification by 23% compared to rule-based systems. The system provides scalable, multilingual, and privacy-preserving health assistance without requiring professional consultation. Future work will integrate wearable health data and emotional health detection.