Automatic Categorization of Educational Videos According to Learning Styles

A learning style consists of a series of personal characteristics with which a person is born and develops as they grow. It determines, among other things, through which activities and senses a person tends to absorb information more efficiently. Learning styles can be used in the process of recommending those Learning Objects (LO) most appropriate to the student's learning style. In this paper, a method to automatically tag videos from a repository of LOs is proposed. The idea is to classify, applying machine learning techniques, a dataset of LOs for the different learning styles with which students are identified. The proposed approach is being evaluated using a repository of videos from the Universitat Politècnica de València.