Research on Online Learning Experience Evaluation Based on LSTM Algorithm and Multimodal Data

• In Event: Advances in Psychometric and Machine Learning Approaches to Assessment

Abstract

This study proposes a multimodal online learning experience measurement model based on Long Short-Term Memory (LSTM) networks to address the subjectivity and time delays of traditional self-report methods. By integrating eye-tracking, EEG, and facial expression data, a deep feature representation framework was constructed. High-precision instruments captured synchronized physiological–behavioral signals, with labeled questionnaire data used for supervision. A sliding window strategy enhanced temporal resolution and model generalizability. The LSTM model achieved 91.66% accuracy, outperforming conventional algorithms. This approach enables real-time monitoring and dynamic feedback, supporting personalized and adaptive learning strategies, and offers a scalable solution for intelligent assessment in online education.

Authors

• yiwei gou, East China Normal University

• Yaofeng Xue, East China Normal University

• Zhan Chen, East China Normal University

• jialin zhou, East China Normal University

• 扣琪 王, East China Normal University