Personalised intelligent multi-agent learning system for engineering courses

Abstract

The paper aims to present a methodology to personalise learning and a model of personalised intelligent multi-agent learning system for engineering courses based on students' learning styles and another personal characteristics and needs. The main technologies used to create the system are Semantic Web, ontologies, recommender system, and intelligent software agents. First of all, the authors performed systematic review on intelligent software agents' application in education. After that, they have analysed students' preferences to certain learning styles according to Felder and Silverman Learning Styles Model which is widely recognised the most suitable for engineering disciplines. This analysis is necessary to further creating personalised learning units / scenarios optimised for particular learners in conformity with their learning styles and other preferences. These learning units should consist of suitable learning components (learning objects, learning activities, and learning environments) optimal for particular students. Scientific methodology to creating optimised learning units for particular learners is based on expert evaluation method and application of intelligent technologies - ontologies, recommender systems, and intelligent software agents. The novel model of personalised intelligent learning system for engineering students based on application of intelligent software agents is presented in more detail. The main success factors of this approach are application of pedagogically sound vocabularies of the learning components used to create personalised learning units, and the experts' collective intelligence.