A demographic-aware hybrid intelligence model for adaptive academic performance prediction in IT-based learning systems

Forecasting student performance correctly in IT-based learning environments is challenging because of the wide range of learner demographics, learning styles, and contextual factors. Traditional machine learning methods frequently regard demographic attributes as static or additional inputs, so limiting their capacity to identify adaptive and cross-group learning patterns. To overcome this limitation, this study introduces a Demographic-Aware Hybrid Intelligence Model (DA-HIM) for the adaptive prediction of academic success in IT-based learning systems. The proposed framework combines data-driven learning analytics with rule-based intelligence to create a dynamic model of how demographic factors interact with behavioral learning features taken from learning management systems. A cross-adaptive approach is presented to modify model parameters across demographic subgroups, facilitating transfer of data while maintaining group-specific learning dynamics. The hybrid architecture uses ensemble machine learning classifiers and an adaptive weighting technique that focuses on fairness, robustness, and predictive stability across different groups of students. Evaluations carried out on real-world IT education datasets show that the proposed DA-HIM framework consistently exceeds traditional single-model and non-demographic baselines when it pertains to prediction accuracy, F1-score, and demographic consistency. Additionally, the adaptive learning technique significantly reduces performance bias among demographic groupings, hence enhancing equitable predictive results. The results show that demographic-aware hybrid intelligence performs well for individualized academic interventions, finding vulnerabilities early, and making decisions based on data in IT learning environments. It provides a scalable and comprehensible method for improving student assistance and educational results in digitally mediated educational environments. DA-HIM integrates demographic attributes with LMS behavioral data to model adaptive and cross-group academic performance patterns. Cross-adaptive hybrid ensemble framework enhances prediction accuracy, robustness, fairness, and stability across diverse student demographics. Adaptive weighting significantly reduces demographic bias, enabling equitable, scalable, and data-driven academic intervention strategies.