Framework for data model to personalized health systems

When large amounts of data is handled, it is important to obtain the desired compatibility between such data to perform activities of access and storage of information; data models are a tool that helps to determine the structure of the information, in order to improve communication and accuracy in applications that use and exchange data with each other for a common purpose. Nowadays, there is no framework for health supporting the data modeling design, i.e. the existing models are generic and therefore are not suitable to support personalized systems and they do not consider the quality of clinical and personal data, required in health care. Based on the CRISP-DM methodology, a framework is proposed to design a data model for personalized health systems. This framework ensures the security of personal and clinical data to relate it with health standards, particularly with the Personal Health (PHR) ISO/TR 14292 standard, which addresses the recommendations of the parameters that must be within a personalized health system. To perform accurate recommendations it is important to make a data mining process, where the data is related to guarantee an accurate and reliable personalization; these relations generated by the model should be taken into account to apply them a data mining technique.