A deep understanding of a balanced diet required by our body for the perfect amount of vitamins and nutrients in order to stay fit and healthy and be in a good shape has been taken up by many researchers who primarily work on building an efficient balanced diet recommender system. Moreover, unlike previous works, the focus of our study is not only on just recommending food items at random but also to offer a list of items keeping in mind one’s taste, choice, and/or preference. Such systems provide a tailor-made diet chart with a perfect balance of health requirements and one’s dietary needs without having to keep visiting health and nutrition experts. In this paper, a diet recommender system is proposed using k-modes clustering and RDF (resource description framework) ontology. Here, based on one’s preference and the calories and other nutrition requirements, the food items are recommended. The developed system takes into account the amount of carbohydrates, fats, proteins, etc. required by any individual as per one’s age and vital statistics. For every cluster, the respective centres are created, and the food items are recommended based on the choice made by the user. At the end, the experimental results are shown where it is inferred that the system performs quite well. Thus, the work contributes to the society by implementing proper health management based on optimally balanced diet recommendations. It is high time to define and implement sustainable approaches towards health management with proper nutritional intake to create an efficient food industry that can aid people in adopting healthy habits by transforming the recommendations into simplified dietary advice expressed in terms of food and food patterns.  

Nutrition; balanced diet; k-modes clustering; machine learning; diet recommender system