Multi-Criteria Decision Analysis Approaches for Selecting and Evaluating Digital Learning Objects

Abstract

The wide spread of internet resources has emerged new paradigms of learning and knowledge delivery to facilitate and enhance learning particularly in e-learning systems. The important milestones of these new paradigms are known as digital learning objects (DLO) which are smallest packed bits for learning. The abundance of these DLOs raises an important question on how to select effectively high quality reusable learning objects. There exist many soft computing approaches such as; fuzzy computing, neural networks, evolutionary computing, support vector machines, machine learning and probabilistic reasoning. In this paper, fuzzy multi-criteria decision-making methods in choosing suitable DLOs were reviewed. This paper discusses recent variety of soft computing methods used in selecting and evaluating digital learning objects through multi criteria decision analysis approaches used; for selecting metrics like basic topical, course similarity, internal topical, basic and user similarity personal, and context similarity situational relevance and for evaluation; scalarization method, employing triangular, trapezoidal and distance based similarity (adapted from TOPSIS techniques). As DLOs continue to evolve it is inevitable to utilize multi-criteria techniques for selecting and or improving quality. The abundance of DLOs has increased the need for applying practical soft computing techniques to retrieve high quality, reusable DLOs.