A Recommendation System for Diabetes Detection and Treatment

Detection, and recommendation systems are widely used in many companies and organizations in favor of giving their employees or even customers a better service. The importance of the systems motivated the investigation of the existing approaches to test its efficiency and possible usage. In this article, the aim is to integrate both systems to create a system that is able to detect diabetes and then recommend a proper plan or medication to overcome diabetes. So, we conducted a study of different approaches of both detection and recommendation systems in order to select the appropriate one. Recommendations systems have different methods. In this paper, we focused on the content-based recommendation method, in which we evaluated using a movie recommender application and movie dataset. We also evaluated and tested four different detection approaches, in which we found that the most accurate approach is the random forest with an accuracy of 79.2% and F-measure of 0.787. A discussion of the usage of a collaborative recommendation system is also introduced to give a wide vision and comparison of the two methods to decide which one is better for a diabetes treatment recommendation system.