Data Flow Framework: A persona-based repository to modeling recommender systems

The present project comes to propose a framework of data acquisition and structured repository in order to be able to apply machine learning techniques to content recommendation. For this, it is intended to use as a basis the modeling and prediction of future scenarios created from the AvaSUS (Virtual Learning Environment of Single Health System in Brazil) data, coverage of primary care in Brazil data, quality indicators from primary care, health facilities and health professional data, as well as data from diseases with mandatory notification. The AvaSUS data are obtained directly from the its database, whereas the others are part of a data lake built for research. The construction of a repository is due to the need of conformation of the structures, to this was chosen to work with the Experience API (or xApi) which defines an ontology structuring users and resources from one action.