Student Modeling

Abstract

This chapter describes how to build student models for intelligent tutors and indicates how knowledge is represented, updated, and used to improve tutor performance. It provides examples of how to represent domain content and describes evaluation methodologies. Several future scenarios for student models are discussed. For example, we envision that student models will support assessment for both formative issues (the degree to which the student has learned how to learn – for the purposes of improving learning capacity and effectiveness) and summative considerations (what is learned– for purposes of accountability and promotion). We envision that student models will track when and how skills were learned and what pedagogies worked best for each learner. Moreover, they will include information on the cultural preferences of learners, their personal interests, learning goals, and personal characteristics. Ultimately, student model servers will separate student models from tutors and will be a part of wide area networks, serving more than one application instance at a time.