Towards Designing AI-Enabled Adaptive Learning Systems

Abstract

Among the many innovations driven by artificial intelligence (AI) are more advanced learning systems known as AI-enabled adaptive learning systems (AI-ALS). AI-ALS are platforms that adapt to the learning strategies of students by modifying the order and difficulty level of learning tasks based on the abilities of students. These systems support adaptive learning, which is the personalization of learning for students in a learning system, such that the system can deal with individual differences in aptitude. AI-ALS are gaining traction due to their ability to deliver learning content and adapt to individual student needs. While the potential and importance of such systems have been well documented, the actual implementation of AI-ALS and other AI-based learning systems in real-world teaching and learning settings has not reached the effectiveness envisaged on the level of theory. Moreover, AI-ALS lack transferable insights and codification of knowledge on their design and development. The reason for this is that many previous studies were experimental. Thus, this dissertation aims to narrow the gap between experimental research and field practice by providing practical design statements that can be implemented in effective AI-ALSs.