Non-Parametric ICA reveals Learning Styles in Education Activities through the Web

This paper presents a new application of independent component analysis to e-learning. An ICA model is proposed defining the sources as dimensions of the learning styles of the students. A novel non-parametric ICA and standard ICA algorithms are applied to huge historical web log data from a virtual campus in order to detect relationship between web activities and learning styles. The data are divided by the course types in: graduate and regular academic career courses and each of those divisions is separated in two subsets: cases with grades and cases with no grades. Web activities include events as course access, email exchange, forum participation, news reading, chats and achievements. Suitable learning styles of the students were positively detected for graduate courses with grades using the non-parametric ICA algorithm.