Sistema de recomendación basado en k-nn para condiciones de incertidumbre en un Sistema Tutor Inteligente.

The development of Intelligent Tutoring Systems aims to create tools to teach a knowledge domain in a similar way as would an experienced human tutor. To achieve it the system should be able to model the state of knowledge and recommend the best suited materials and interventions for each student. However, the knowledge associated with how people learn, the most effective ways to transmit knowledge and to measure it is incomplete and error prone. For these reasons, student modeling must take into account the uncertainty inherent in the process. This paper proposes a technique for developing a student module with the use of Bayesian networks and collaborative filtering, linking the stages through a measure of effectiveness generated during the assessment, called coefficient of effectiveness. [ABSTRACT FROM AUTHOR]