This study addresses the high dropout rates in online STEM education by developing a novel dropout prediction model using MOOC learning behavior data. The model integrates convolutional and recurrent neural networks with a long short-term memory mechanism to predict dropouts based on explicit and implicit features of learning behavior. The study analyzes key temporal sequences of the learning process, identifies impacting factors, and deduces intervention measures. The results show high prediction accuracy and provide effective strategies for improving STEM learning effectiveness in MOOCs.
Publisher
Humanities & Social Sciences Communications
Published On
Mar 18, 2024
Authors
Xiaona Xia, Wanxue Qi
Tags
dropout prediction
online STEM education
MOOC
neural networks
learning behavior
intervention measures
prediction accuracy
Related Publications
Explore these studies to deepen your understanding of the subject.
