Framework para predições e recomendações em dados acadêmicos

Abstract

In virtual environments, records of student productions, activities, and interactions are collected automatically, enabling analysis of development and learning. With this, it is possible to identify patterns of how students learn. Therefore, favoring the appropriate basis for the appropriate recommendations. The goal is to present a framework for generating messages of recommendations based on predictive systems in an innovative way. To validate the work was performed the analysis of a public data set.