A Recommender System for Weight Loss Activity applying Ontological Knowledge-Based

Abstract

Nowadays, the behavior of human life has been changed continuously and many people encounter the problems about health especially in fat reducing. This research presents how knowledge-based combined with rules can be used to develop the recommender system for weight loss activity to help the people for recommendations and produce a fat reducing plan. The question is how to automatically create a suitable plan and recommendation for weight loss activity, because the plan and recommendation should be compatible with individual people. Thus, the knowledge-based and rules in this research are proposed to solve the problems. There are three objectives of this research including 1) creating the ontology to represent the conceptual knowledge-based for activities of fat reducing and related information such as fat reducing activities, food, and disease, 2) creating the rules combined with proposed knowledge-based, and 3) developing the prototype system using proposed knowledge-based and rules to evaluate this approach. The accuracy is in the good level (average score = 0.73 (73%)). It is calculated using accuracy metrics by predictions with real users’ preferences and domain experts from real environment, 2 months.