Development knowledge management for adaptive hypermedia learning system

Abstract

Adaptive hypermedia learning systems (AHLS) aims in adaptively accommodating learning materials based on individual differences in user. This kind of system faces problems in cognitive overload where too many information are given to students resulting to some of them become lost in hyperspace. In order to reduce the overload, user or leaner should be given materials well suited with their learning ability and also in their style of learning. Our project concentrates in individualizing the learning material and navigational paths to adapt with different users learning styles and knowledge acquisition. Learning ability is associated by their knowledge acquisition in using the system captured via user model. Knowledge acquisition per se is not enough as research on learning shows students learns effectively when they were thought with methods that suits their learning style. Pedagogical framework for this project comprise of Myers Briggs Type Indicator (MBTI) personality factor as learning strategy and Howard Gardner and Honey & Mumford theory for learning technique and method. Our contribution is on the learning materials which are structured based on the pedagogical framework. Another contribution is mainly on the theoretical and practical mechanism for the integration of computational intelligence techniques mainly to personalize the user both at the presentation and navigation level. Theoritical and practical aspects of this project are discussed; e.g., instructional material suitable for learning style chosen, experiments on Kohonen neural network and rough sets for classification of knowledge acqusitions, fuzzy logic architecture for the adaptation in learning style, architecture of the prototype developed and also its evaluation. The contribution of this project is based on these seven papers describe herewith: Paper I discusses the problems in personalizing instructional material that matches the differences of students according to student’s fuzzy membership to certain learning styles. Paper II focuses on the use of computational intelligence techniques such as Kohonen selforganizing maps in the classification of student models and Fuzzy logic in the adaptation of learning material and navigation path. Paper III discusses the structure of the learning material which are; 1) theory, 2) example, 3) exercise and 4) activities. In paper IV, a generation of rough set rules is implemented in identifying the status of students’ knowledge acquisition in hypermedia learning. Paper V describes the use of back propagation neural network in classifying the student. Paper VI presents the architecture and design for the development of an adaptive hypermedia learning system for teaching and learning Data Structure. Paper VII discusses the evaluation of an adaptive hypermedia learning system developed. We use summative evaluation where we address the educational impact of this system on students and its practical acceptability in terms of usability.