Static and dynamic user type identification in adaptive e-learning with unsupervised methods

A key factor in modern e-learning systems is the correct identification of the user learning style, to provide appropriate content presentation to each individual user. Moreover, a continuous user monitoring is essential in assessing the progress made during the learning process and controlling the desired evolution. In this paper we present a strategy for integrating the static and the dynamic user models, in a previously proposed e-learning system. Also, we assess the static user models through unsupervised learning techniques and establish that a 3-type model is more appropriate, validating previous analyses performed by a domain expert.