Representasi Pengetahuan pada Web Semantik untuk Sistem Rekomendasi Menu Diet Berbasis Ontologi

Abstract

Unhealthy dietary habits and generalized nutritional recommendations remain major contributors to chronic diseases, despite increasing public awareness of healthy lifestyles. To address the limitations of conventional diet recommendation systems, this research proposes an ontology-based diet menu recommendation system built on Semantic Web principles. The system utilizes structured knowledge representation using ontologies to model food, nutrients, diet types, and user preferences. It employs a hybrid filtering method combining content-based and collaborative filtering to generate contextual and personalized recommendations. The ontology was developed using the Methontology framework and implemented in Protégé, while the system was built using Next.js and integrated with Apache Jena Fuseki for SPARQL-based reasoning. System evaluation covers ontology verification and validation, functional testing via black-box method, recommendation performance using Precision@10, and user acceptance measured by the Technology Acceptance Model (TAM). Results show high accuracy in recommendations (average Precision@10 = 0.72), and users found the system both useful and easy to use, validating the effectiveness of semantic technologies in personalized diet planning.