Adaptive Cloud-Integrated Artificial Intelligence for Personalized Learning Pathways in Higher Education

Abstract: At a time when Higher Education (HE) institutions are struggling to improve student engagement and learning outcomes while closing achievement gaps within their increasingly diverse student communities, personalized learning (PL) in the form of tailored learning pathways is gaining traction. However, the cost, resource, and time implications in the design of PL solutions, especially in Higher Education, have constrained broader PL uptake, initial attempts in so- called personalized artificial intelligence (AI) have neglected critical dimensions of the student modelling component, and adaptation to actual student needs has remained an aspirational goal rather than reality. By drawing on innovative concepts from learning analytics, education data mining, adaptive instructional design, and adaptive cloud-integrated AI research, an adaptive cloud-integrated AI system is proposed that empowers any educator to create a PL solution in their discipline yet adapts to evolving student needs and interests in real-time, requires no up-front configuration, and can be deployed across disparate subject areas and learning contexts.

The architecture encompasses a cloud-deployed data foundation that supports student modelling and profiling, a recommender engine, and an adaptive personalization engine. Elements and algorithms within the three system components are presented and showcased in a range of Higher Education disciplines, illustrating how the integration of existing student activity data with cloud-hosted Repository and Knowledge Graph data learning pathways can be designed to mirror unit, course and program-level Learning Outcomes antennas without the time, resourcing or ongoing expertise overhead of traditional solutions. Insights from Greater Sydney and Auckland case studies indicate that PL, in various forms, improves student engagement and perception of learning outcomes. Disparate student characteristics add support for a degree of equity and access inference. However, the personalization process must be customized and continually refined to maximize its catering for the full range of student diversity.

Keywords: Personalized Learning in Higher Education, Adaptive Learning Pathways, Cloud-Integrated Educational AI, Student Modelling and Profiling, Learning Analytics, Education Data Mining, Adaptive Instructional Design, AI-Driven Personalization Engines, Recommender Systems for Learning, Knowledge Graph–Based Learning Design, Repository- Integrated Learning Content, Real-Time Learning Adaptation, Learning Outcomes Alignment, Scalable Personalized Learning Architectures, Equity and Access in Education, Student Engagement Analytics, Adaptive Educational Systems, Cross-Disciplinary Learning Platforms, Cloud-Based Learning Infrastructure, Intelligent Higher Education Systems.