Information and communication technologies make it possible to bridge the gap and time barriers in the flow of health information and knowledge, allowing every involved part in the health process to have access to the information. This approach provides the knowledge of the individual to contribute effectively to the improvement in human health. But also, helps the collective knowledge effectively to solve health problems on individual level. In this paper we are evaluating the algorithm that generates recommendation for users. We are using simulations on generic data to see how different types of activities are affecting the accuracy of the algorithm. On the basis of the performed activities and blood glucose measurements, our recommendation algorithm should determine list of activities that have bigger influence on the change of the blood glucose levels. Generic data for our simulations are based on modeling of food intake and physical activity influence over the blood glucose level. © Springer International Publishing Switzerland 2015.