NutriTrack: A Health-Aware Food Recommendation System for Diabetes and Hypertension Using Rule-Based Filtering and Graph-Based Analysis

Lifestyle-related diseases such as diabetes and hypertension are increasing rapidly. This has increased the importance of maintaining healthy and balanced eating habits. However, preparing personalised meal plans is difficult because nutritional requirements differ from person to person. Factors such as age, body weight, activity level, food preferences, and medical conditions must all be considered carefully. This paper presents NutriTrack, a health-aware food recommendation system designed to generate personalised daily meal plans. Nutritional information and user-specific factors are used to produce appropriate meal recommendations. NutriTrack combines rule-based filtering with multi-nutrient analysis to create balanced meal plans tailored to individual requirements. Recommendations are generated by considering calorie requirements, dietary preferences, carbohydrate intake, protein content, fibre intake, sugar limits, and sodium levels. Additional dietary constraints are incorporated for users with diabetes and hypertension to improve the suitability of the recommended meals for their health conditions. An experimental Graph Convolutional Network (GCN) model was evaluated to examine nutritional relationships between food items based on their nutritional characteristics. The model achieved an accuracy of 86.50 The findings show that NutriTrack can provide meal recommendations that match different dietary needs, including for people with health conditions such as diabetes and hypertension. The system also focuses on being transparent about how recommendations are made, helping users understand their food choices and make better decisions in their daily lives.