A neurofuzzy-based quality of eLearning model

Abstract

Higher education institutions are increasingly implementing eLearning means to meet the increasing demand of students' enrolment. Current eLearning quality models have been reviewed. A two-stage softcomputing eLearning quality model, SCeQLM, is proposed. The SCeQLM model is based on ten critical success factors (CSFs). Characteristics of each CSF are modelled, the outputs input into the second stage, to produce the overall output of the SCeLQM model. This output will be either low, satisfactory, good or high. To validate the adequacy of these models, correlation coefficient and mean absolute percentage error (MAPE) have been used. We achieved correlation coefficients values of higher than 0.9905 and MAPE values lower than 1.722 for all the models. The promising results indicate the relevance to apply the modelling techniques, neurofuzzy, for this type of problems. As a future work, our goal will be to further investigate, enhance and develop a web-based SCeLQM version.