A Systematic Review of Educational Recommender Systems: Techniques, Target Users, and Emerging Trends in Personalized Learning

Abstract

This systematic review investigates the state of educational recommender systems (ERS) to synthesize trends, techniques, user focus, and research gaps in the domain. Guided by the PRISMA (Page et al., 2021) checklist, relevant studies were selected and analyzed based on key elements such as the recommendation techniques employed, target user groups, and application contexts. The thematic synthesis revealed that machine learning remains the most widely adopted approach, particularly classifiers, clustering, and ensemble methods. Collaborative filtering, hybrid models, and ontology-based approaches featured prominently, though techniques such as deep learning and genetic algorithms were underutilized despite their potential. Most systems primarily targeted students, with relatively limited attention given to educators, administrators, and lifelong learners. Application areas such as course selection, career guidance, curriculum development, and digital literacy support were observed. However, challenges such as limited inclusivity, contextual adaptability, and the underrepresentation of workplace learners were noted. The findings underscore the need for more inclusive, scalable, and context-aware recommender systems. The review highlights opportunities for advancing digital literacy, helping users to navigate through digital tools, especially in the workplace and non-formal learning environments, and calls for future research to explore ethical, explainable, and cross-regionally deployable ERS frameworks. The study contributes a comprehensive synthesis that informs both the academic discourse and practical development of personalized, user-centered educational recommender systems.