Cloud eLearning - A New Paradigm of eLearning System Using Artificial Intelligence Techniques for Generating Personalized Learning Paths

Background/purpose. With advancements in technology, particularly in Artificial Intelligence (AI), personalized and adaptive systems are increasingly being integrated into conventional educational environments. These technologies create opportunities to place learners at the center of the educational experience through personalized learning. However, the lack of standardization among online educational learning resources limits their adaptability and re-use across different systems and contexts. To address this challenge, this study introduces a novel idea for an adaptable artificial intelligence system, namely the Cloud eLearning system (CeL), which includes Cloud eLearning Learning Objects (CeLLO) —standardized learning objects tagged with semantic metadata for flexible re-use. Materials/methods. The system architecture features a Cloud eLearning Recommender System, which ranks educational learning resources using a vector space model and hierarchical clustering, and a Cloud eLearning Automated Planner, which generates personalized learning paths based on the recommender output. The study employed a two-phase evaluation methodology. First, the functionality of the AI-based Cloud eLearning prototype was assessed. Second, a user evaluation was conducted through a questionnaire administered to 17 undergraduate and postgraduate students selected from five Computer Science courses. The questionnaire collected both quantitative and qualitative data, including learner satisfaction and perceptions of personalization services. Results. Quantitative data were analyzed through correlation analysis, revealing significant correlations: 99% between system performance and personalization services, 99% between system performance and learner satisfaction, 90% between system performance and learners’ initial background, 95% between learners’ background and personalization services, and 99% between learner satisfaction and personalization services. Conclusion. These findings highlight the effectiveness of CeL in delivering personalized learning experiences and suggest that future work should focus on further validating the system across broader educational contexts and refining the algorithms output.