A MULTI AGENT FRAMEWORK FOR EFFECTIVE CHANGE MANAGEMENT OF LEARNING CONTENTS IN SEMANTIC E-LEARNING

 

Sohail Sarwar¹, Zia Ul Qayyum¹, M Safyan², Muddessar Iqbal¹

¹ University of Gujrat Gujrat, Pakistan, sohail.sarwar@seecs.edu.pk, zia.qayyum@uog.edu.pk

²GC University Lahore, Pakistan, m.safyan@seecs.edu.pk

 

 

Abstract

The inevitable changes in learning contents need to be tracked and managed seamlessly for knowledge repositories in order to facilitate the learners with up-to-date concepts. However, evaluation of suggested changes (in lessons, topics, sub-topics, learning objects etc) for incorporation into active knowledge base is of key importance identified through different factors such as content change type, change module, change priority, learning content change validity etc. A framework based on intelligent agents has been proposed for identifying/incorporating valid changes in the repositories of learning contents (and learning objects) on the basis of given parameters through different agents. These agents have been employed for different tasks such as Content Change Receptor Agent (CCRA)” to gather change specific parameters for use by “Learning Object Change Analyzer Agent (LOCAA), “Manage Learning Repository Agent (MLRA)” for handling the tasks specific to semantic repositories and “Learning Content Change Revision Agent (LCCRA)” to intelligently predict a change (through knowledge engineering techniques) if given parameters don’t recognize the change in learning content(s). Moreover, these agents are intelligent enough to interoperate with latest generations of web i.e. web 3.0 (semantic web). This interoperation of agents is enabled through different baseline ontologies devised for modeling the learning contents in the form of lessons, topics, sub-topics their annotations and instances. In order to evaluate the proposed framework, a twofold approach has been exploited i.e. evaluation of ontological models and performance accuracy of system agents for catering the change in learning contents. Here, it is worth mentioning that initial implementation and evaluation parameters of proposed models are envisaged to be in controlled environments with evaluation in real time academic conditions over the years.

Keywords: E-Learning, Semantics, Multi-Agents, Knowledge Engineering, Learning Objects