Published Paper Details:

This study delves into how student engagement with Blackboard, a prevalent Learning Management System in higher education, shapes their academic achievements. Employing a mixed-methods approach, it examines four deep learning models to predict student performance using Key Performance Indicators (KPIs) derived from Blackboard data. Analyzing data from seven courses, potential predictive KPIs are identified through documentary analysis. Correlational studies unveil significant associations between these factors and student performance metrics. The study demonstrates the efficacy of a combined CNN-LSTM predictive model in accurately forecasting outcomes. These findings advocate for the utilization of such models to optimize Blackboard's utility and bolster educational interventions in universities. Additionally, the integration of BI-LSTM extends the study's accuracy range to 95-100%, showcasing its potential in refining predictive capabilities further.