3D-BCLAM: A Lightweight Neurodynamic Model for Assessing Student Learning Effectiveness.

Area of Science:

• Educational Technology
• Cognitive Science
• Artificial Intelligence

Background:

• Evaluating student learning effectiveness is crucial for understanding learning processes and barriers.
• Emotional factors significantly influence learning outcomes, offering a new perspective for assessment.
• Traditional evaluation methods have limitations in feature extraction and model complexity, hindering the use of emotional data.

Purpose of the Study:

• To propose an innovative, lightweight neurodynamic model for comprehensive student learning effectiveness assessment.
• To address the limitations of traditional methods in capturing and analyzing emotional data during learning.

Main Methods:

• Developed a novel model named 3D-BCLAM, integrating Bidirectional Convolutional Long Short-Term Memory (BCL) and a dynamic attention mechanism.
• Designed the model for efficient capture of emotional dynamic changes in time series data with low computational cost.
• Evaluated the model's performance on public datasets.

Main Results:

• 3D-BCLAM demonstrated outstanding performance in evaluating student learning outcomes.
• The model significantly outperformed traditional machine learning and deep learning models (CNN, RNN).
• The model achieved comprehensive assessment, covering both cognitive and emotional dimensions.

Conclusions:

• The 3D-BCLAM model offers an effective and computationally efficient approach to learning effectiveness assessment.
• Integrating emotional dynamics enhances the depth and accuracy of student performance evaluation.
• This work supports innovation in student learning assessment by leveraging emotional data.