A Computational Architecture for Learning Analytics in Game-Based Learning

Abstract

The use of digital games in education is already a reality, but their use in this area has been hampered by the lack of consolidated resources for evaluation of game-based learning. This paper reports a computational architecture for learning analytics in game-based learning that is based on relational analysis and data mining of data containing evidences of learning collected during the game play. It also uses computer vision techniques applied on images of the players playing to obtain information about their behaviors and emotions during the game.