Recommending Healthy Food for Phenylketonuria with Collaborative and Content-Based Filtering

Phenylketonuria (PKU) is a genetic disorder requiring strict dietary management to prevent intellectual disability. Existing food recommendation systems struggle to address the unique needs of PKU patients, often neglecting the importance of balanced nutrition alongside protein control. This paper proposes a novel, age-based recommender system for PKU patients. The system leverages ontologies to capture knowledge about PKU-related knowledge, foods, and nutrients, ensuring safe and personalized meal recommendations. It personalizes suggestions by considering a patient’s PKU diagnosis, protein tolerance level, and potentially their preferences. For infants, a rule-based system ensures safe protein consumption and suggests suitable recipes. Adults benefit from both content-based and collaborative filtering techniques. This work has the potential to empower PKU patients to effectively manage their dietary needs by providing safe, personalized, and diverse meal recommendations.