The Minimum Normalized Dissimilarity between Objects based Rough Set Technique for Elucidating Learning Styles in E-learning

Abstract

Objectives: The objective of this research work is to analyse the learning styles of individual users in an e-learning system and to formulate a mathematical model to determine it. Methods: This research work proposes MNDBO, a rough set based clustering technique for elucidating learning styles by finding minimum normalized dissimilarity between objects in e-learning. The proposed clustering technique uses a normalized score value for estimating the deviation between data’s through the equivalence property of rough set theory. Findings: Further, the result predicts that the clusters produced by MNDBO algorithm perform better than MADO by 11%, 14% than SDR and 16% than MMR in terms of cohesion. Furthermore, MNDBO algorithm also produces better results than MADO by 15%, 23% than SDR and 27% than MMR in terms of coupling. In addition MNDBO algorithm maximizes the cohesion and simultaneously reduces the coupling rate based on varying number of cluster size on an average 15% and 19% respectively. Applications/Improvements: If this Rough set based clustering technique is used means we can able to discover successfully relations with inconsistent or incomplete data.