Online Learning Behavior Analysis and Prediction Based on Spiking Neural Networks

Abstract

The vast amount of data generated by large-scale open online course platforms provide a solid foundation for the analysis of learning behavior in the field of education. This study utilizes the historical and final learning behavior data of over 300 000 learners from 17 courses offered on the edX platform by Harvard University and the Massachusetts Institute of Technology during the 2012–2013 academic year. We have developed a spike neural network to predict learning outcomes, and analyzed the correlation between learning behavior and outcomes, aiming to identify key learning behaviors that significantly impact these outcomes. Our goal is to monitor learning progress, provide targeted references for evaluating and improving learning effectiveness, and implement intervention measures promptly. Experimental results demonstrate that the prediction model based on online learning behavior using spiking neural network achieves an impressive accuracy of 99.80%. The learning behaviors that predominantly affect learning effectiveness are found to be students’ academic performance and level of participation.

Keywords

• online learning
• learning outcomes prediction
• learning behavior analysis
• spiking neural network