On top-down versus bottom-up personalisation and evaluation of augmented reality learning systems

Abstract

The paper aims to analyse the problem of personalisation and evaluation of quality of Augmented Reality (AR) learning systems. AR is often used in education to enhance students’ motivation by visualizing learning content and activities. In the paper, first of all, systematic review of relevant scientific literature on the research topics was conducted. The author’s AR learning systems quality personalisation and evaluation frameworks are presented in the paper. Evaluation of quality of AR learning systems should be based on applying both expert-centred (top-down) and user-centred (bottom-up) quality evaluation methods consisting of creating quality models (systems of criteria) and evaluation methods. AR-based learning systems including learning content (i.e. learning objects) and activities should be suitable, acceptable and usable for particular learners. Personalisation of AR learning systems should be based on learners’ models/profiles using students’ learning styles (bottom-up method), and educational data mining (top-down method). AR personalisation method is aimed to personalise learning by applying well-known learning styles models, educational data mining methods and techniques, and intelligent technologies, and thus to ensure that suitable AR-based learning systems should be selected for particular users to improve their learning motivation and thus—quality and efficiency. The method of identifying students preferring to actively use AR-based learning systems is based on identification of probabilistic suitability indexes to choose the most suitable AR-based learning systems for particular students. Experimental research is also performed, and its results are presented in the paper. The research is multidisciplinary, including computer science, education, operations research, and educational psychology areas.