Abstract: This paper demonstrates the framework of a system built to monitor health records for diabetes patients and extracting data on a frequent basis to objectively predict the impending health risks using a version of Bayesian network. The system is deployed on a SaaS cloud service to facilitate secure and convenient collection of data from patients using a mobile application and a web console.

Keywords: Diabetes Mellitus, Data mining, Bayesnet, Support vector, patient monitoring, HER.