On the classification of interactive user behaviour indices

ABSTRACT

The concepts of user behaviour entropy and user behaviour mobility are proposed as indices for the description of user behaviour. The user behaviour indices are derivable from the mode probability vector and the mode transition matrix which adequately describe the behaviour dynamics of an interactive user. The user behaviour indices reduce the ((n*n)+n) dimensional parameter space to two dimensions only for classification, without loss of information related to the user behaviour dynamics. The classification of the users in an interactive educational environment using the user behaviour indices is presented as a case study.

Index Terms

• Computing methodologies

  • Machine learning

    • Learning paradigms

      • Supervised learning

        • Supervised learning by classification

    • Machine learning approaches

      • Classification and regression trees

• Human-centered computing

  • Human computer interaction (HCI)

    • Interaction paradigms