Application of aggregated similarity measures in evaluation of e-learning web sites
Milan Mrkalj · Info M · 2011
Main goals of this paper are to define the concept of quality of sites for e-learning and importance of its evaluation, and to increase the ability of good, objective, and comprehensive evaluation through mathematical modeling of similarity measures. Similarity measure modeling is based on a new paradigm that stems from the theoretical concepts of Interpolative Boolean Algebra (IBA), which advances the concept of Case-Based Reasoning (CBR) and uses Interpolative Boolean Algebra (IBA) and Logical Aggregation (LA) as aggregation operators of similarity measures, in order to enable secondary (relative) quality evaluation of e-learning websites. Secondary evaluation is done on the basis of the degree of similarity with some of the 'prototype' sites that are ranked (ordered) by experts in the primary (absolute) evaluation. The comparative analysis is shown for different types of realization of similarity measures modeling (Logical Aggregation, Euclidean norm, and Artificial Neural Networks).