DEVELOPMENT OF WEB-BASED ADAPTIVE LEARNING RESOURCES SEQUENCING MODEL

Abstract

In recent times, there is massive acquisition of knowledge through the use of computing and electronic technologies to access learning resources outside traditional classroom environment, which has greatly increased educational accessibility globally. As a result of this interesting development, several hundred integrated Learning Management Systems (LMSs) offer different approaches to supporting the growing e-learning field have emerged. Therefore, Instructors (teachers) have access to use any of these LMSs for the creation of courses and assessment suites, for communicating with the students, for monitoring and evaluating their work; students can also learn, communicate and collaborate with others. However, LMSs do not offer personalized services, all the students are provided with the same set of educational resources and tools, without taking into account the differences in knowledge level, interests, motivation and goals. Therefore, in this thesis, an adaptive personalized e-learning is developed based on Felder Silverman Learning Style Model (FSLSM). The learning contents are sequenced by adopting knowledge tree technique where the nodes represent the concepts and the links represent the relationship among the concepts. The architecture of the adaptive system is made up of creation and management of learning object such as the domain modeling using the learning object database; the generation of a personalized learning for the learner or group of learners; the presentation of learning objects to the learner and finally, the prediction of learning style preferences for each learner. The Learner’s knowledge which represents the learner’s interests on visited learning objects with respect to learners educational preference from the first access till the last is computed. k-means clustering is used to divide the learning preferences into eight groups and multiple correspondence analysis is used to partition the learning preference into four dimensions to assign learners with common preferences and interests to the same group. In order to determine the overall performance of the learner on a course at the end of a particular semester, various parameters such as study performance, review performance and assessments are used. The system was implemented using My Structured Query Language relational database management system as backend engine, Hypertext Pre-Processor as frontend engine and Apache as the web server. The chi-square analysis of the comparison between the model and practical (real life test) revealed that there is no significant difference between the learning suites of the model except the suite sequential and global which showed a significant difference with a chi-square value of 30.243 and p-value 0.000 less than five percent significant p-value of 0.005. Finally, the implementation of the FSLSM-based adaptive e-learning system maximizes the subjective learner satisfaction, the learning speed (efficiency) and the assessment results (effectiveness).