A Three-Tier Profiling Framework for Adaptive e-Learning

Abstract

Existing methods support adaptive e-learning mainly by setting student characteristics in a student profile, and use it as a filter to extract suitable learning content from a dedicated structure of course materials. If simple student characteristics, such as prior knowledge and learning preference, are considered, it may be straightforward for an instructor to set up the student profiles. However, if complicated student characteristics, such as learning styles, interaction styles and content styles, and other factors that affect the students' interests on the course materials are involved, it may become too difficult for an instructor to design a suitable course structure matching all these criteria. It is also complicated for system implementation as many rules need to be set up. In this paper, we propose a three-tier profiling framework in conjunction with a concept space structure and a set of concept filters to address the above problems. The framework offers a unified way to model and handle a variety of student learning needs and the different factors that affect course material relevance. The framework is extensible in nature and can form the foundation for the future development of adaptive e-learning systems.