A Hybrid Ensemble Learning and Case-Based Reasoning Framework for Learner Intelligence Prediction and Personalized Adaptive Learning
DOI:
https://doi.org/10.70917/ijcisim-2026-3025Keywords:
Adaptive Learning, Educational Data Mining, Learner Intelligence Prediction, Ensemble Learning, Personalized Learning, Machine LearningAbstract
Due to the growing need for intelligent learning systems that can cater to the different demands of learners, personalized adaptive learning is an essential research topic in educational data mining. Most existing learner prediction models are based on academic achievement and ignore the latent learner traits and tailored recommendation mechanisms. To overcome these constraints, this paper introduces a Hybrid Ensemble Learning and Case-Based Reasoning Framework for Learner Intelligence Prediction and Personalized Adaptive Learning. The proposed system combines Feature Engineering, Principal Component Analysis (PCA), K-Means clustering, Hybrid Ensemble Learning and Case-Based Reasoning (CBR) in one adaptive learning architecture. The learner intelligence is estimated and the learners are classified into five adaptive learning groups using four manufactured learner attributes, Study Habits Score, Academic Participation Score, Academic Performance Score and Learner IQ Score. To examine the robustness and generalization power of the proposed framework in diverse educational contexts two benchmark educational datasets OULAD and xAPI-Edu-Data were used. Experimental findings indicate that the Hybrid Ensemble model obtained 97.92% classification accuracy on the xAPI dataset and still performed well on the OULAD dataset. Furthermore, the CBR module attained an average similarity score of 83.01%, a retrieval accuracy of 99.58%, and a recommendation coverage of 100%, and produced five individualized recommendation methods for distinct learner profiles. Comparative research in both datasets validates the scalability, robustness and efficacy of the proposed framework of learner intelligence prediction and customized adaptive learning. The suggested framework offers an intelligent decision support system that may improve student engagement, educational results, and data-driven adaptive learning environments.
