A modelling framework for constructivist learning in exploratory learning environments

Exploratory learning has been proved to be beneficial for learning when guidance and support is provided. This is a challenging issue as balance between freedom and control is a strong requirement in this context. In this
paper the challenges involved are discussed together with the way they affect the process of learner modelling. We discuss previous attempts to model the learner in exploratory learning, and how they fall short, and propose as an
alternative a learner modelling framework for exploratory learning environments that adopts the constructionist view of learning. We investigate
the application of this framework in the domain of mathematical generalisation
and discuss how the model can potentially support feedback generation. An
example is discussed in detail, and an approach based on case-based reasoning
and soft computing is presented to represent and process short-term and long-
term learner behaviour and knowledge in the learner model.