Unapređenje sistema za e-učenje semantičkom nadgradnjom

Abstract

Within this doctoral dissertation an enhancement to e-learning systems has been proposed and the evaluation of its impact in the context of real learning was performed. The proposed enhancement is based on coupling a notion-relation graph, in the RDF format, with the textual learning material, to enable semantic queries over the graph by dragging and droping the words in the text, in cases when the relation between two notions is unclear and still required for the student to continue learning, with the aim of reducing the need for notion definition searches (distractions). To evaluate the poposed enhancement efficiency an e-learning system simulation, learning text and questions were developed. 191 Faculty of Electronic Engineering students took part in the main experiment and 27 took part in the verification experiment that followed. The key results proved the basic hypotheses: the proposed enhancement significantly reduced the participants' need for searching within the text and their completion time, while it raised their preparedness to proceed with learning. Additional results gave insight into other effects of the proposed enhancement and will be used as guidelines for further development. All key results were confirmed by the verification experiment.