Conceptual change modeling using dynamic bayesian network

Abstract

Modeling the process of conceptual change in scientific inquiry learning environments involves uncertainty inherent in inferring learner's mental models. INQPRO, an intelligent scientific inquiry exploratory learning environment, refers to a probabilistic learner model aims at modeling conceptual change through the interactions with INQPRO Graphical User Interface (GUI) and Intelligent Pedagogical Agent. In this article, we first discuss how conceptual change framework can be integrated into scientific inquiry learning environment. Secondly, we discuss the identification and categorization of conceptual change and learner properties to be modeled. Thirdly, how to construct the INQPRO learner model that employs Dynamic Bayesian networks (DBN) to compute a temporal probabilistic assessment of learner's properties that vary over time: awareness of current belief, cognitive conflict, conflict resolution, and ability to accommodate to new knowledge. Towards the end of this article, a sample assessment of the proposed DBN is illustrated through a revisit of the INQPRO Scenario interface. © Springer-Verlag Berlin Heidelberg 2006.