Discovering Student Preferences in E-Learning Cristina Carmona 1 , Gladys Castillo 2 , Eva Millán 1 1 Departamento de Lenguajes y Ciencias de la Computación, Universidad de Málaga, Spain {cristina,eva}@lcc.uma.es 2 Department of Mathematics, University of Aveiro, Portugal. gladys@mat.ua.pt Abstract. Nowadays modeling user’s preferences is one of the most challenging tasks in e-learning systems that deal with large volumes of information. The growth of on-line educational resources including encyclopaedias, repositories, etc., has made it crucial to “filter” or “sort” the information shown to the student, so that he/she can make a better use of it. To find out the student’s preferences, a commonly used approach is to implement a decision model that matches some relevant characteristics of the learning resources with the student’s learning style. The rules that compose the decision model are, in general, deterministic by nature and never change over time. In this paper, we propose to use adaptive machine learning algorithms to learn about the student’s preferences over time. First we use all the background knowledge available about a particular student to build an initial decision model based on learning styles. This model can then be fine-tuned with the data generated by the student’s interactions with the system in order to reflect more accurately his/her current preferences. 1 Introduction Student modeling is the process whereby an adaptive learning system creates and updates a student model by collecting data from several sources implicitly (observing user’s behaviour) or explicitly (requesting directly from the user). Traditionally, most of student modeling systems have been limited to maintain assumptions related with student’s knowledge (acquired during evaluation activities) not paying too much attention to student’s preferences. However, over the last years the growth of on-line educational data (encyclopaedias, repositories of learning resources, etc.) has made it necessary to “filter” or “sort” the information shown to the student, so he/she can make a better use of it. Since one of the first works in e-learning that suggested the use of learning styles for determining the student’s preferences regarding multimedia materials [1], this research direction has been getting more and more attention. Learning styles can be defined as the different ways a person collects, processes and organizes information. It is a fact that different people learn differently: some people tend to learn by doing, whereas others tend to learn concepts; some of them like better written text and/or spoken explanations, whereas others prefer learning by visual information (pictures, diagrams, etc). On the other hand, different learning resources can explain the same concept by implementing different learning activities Proceedings of the International Workshop on Applying Data Mining in e-Learning 2007