Multi-Attribute Utility Theory and Adaptive Techniques for Intelligent Web-based Educational Software

Abstract

This paper describes how the Multi-Attribute Utility Theory can be combined with adaptive techniques to improve individualised teaching in an Intelligent Learning Environment (ILE). The ILE is called Web F-SMILE, it operates over the Web and is meant to help novice users learn basic skills of computer use. Tutoring is dynamically adapted to the individual learner based on the learner modelling component of the system and the Multi-Attribute Utility Theory (MAUT) that is employed to process the information about the user. As a result, MAUT provides a way for the system to select on the fly the best possible advice to be presented to users. Advice is dynamically formed based on adaptive presentation techniques, where adaptation is performed at the content level and adaptive navigation support, which is performed at the link level of the hyperspace of the tutoring system. The adaptivity of learning depends on factors such as the learner’s habits, prior knowledge and skills, which are used as criteria for the application of MAUT in the educational software. In this way, a novel combination of MAUT with adaptive techniques is used for intelligent web-based tutoring.