Ontology and semantic matching for diabetic food recommendations

Foods recommendation for diabetes patients is indispensable for controlling blood sugar levels. Currently, the foods preparation is done by a nutrition expert. The patient's dependence on the nutrition experts is very high, thus the selection of foods could not be done independently. The Automation system to determine foods combination for diabetic patients is needed to solve these problems. In this study, the automation system has been designed and implemented. The technologies used in this research are the OWL and SWRL. There are few researches that explore an automation process of foods recommendation for diabetes patients using the technology of OWL and SWRL. Domain knowledge based on Ontology is needed to process foods composition automatically. However, using SWRL and OWL technology is not enough, because the accuracy of the words required. A semantic ontology understanding was added using weighted tree similarity method to specify the composition of foods for diabetic patients. 73% data were able to be correctly predicted by this method.