SmartNutri: A Mental Health-Aware Knowledge Model for Personalized Food Recommendation Systems

Abstract

Mental health can influence appetite and personal food preferences. These factors also affect energy intake, medication-related nutritional needs, and the ability to maintain healthy eating patterns. Existing food recommender systems primarily optimize nutritional value or user preferences for physical health conditions, but rarely integrate the mental health context. This study aimed to develop a knowledge model for a personalized food recommendation system designed to support people with mental health conditions. The proposed model integrates user profiles such as body height and weight, self-reported mental health conditions, and food preferences. A combination of the Knowledge Management Cycle (KMC) and Rapid Application Development (RAD) research methodology is used. A resulting knowledge model represents relationships among nutrients, mental health conditions, and demographics, allowing transparent and context-sensitive recommendations. The knowledge model is then used to feed recommendations to the web application, which is the result of RAD. The performance of the proposed system is assessed through combination generation testing and the explainability of the recommendations generated. The proposed approach offers a reusable knowledge representation that bridges nutrition science, mental health considerations, and the design of personalized food recommendation systems. The model is intended to increase awareness about nutrition that is tailored to individual needs rather than replacing clinical care. In a specific setting, the SmartNutri web application can be a tool for health workers to provide food recommendations during counseling in mental health facilities.