K12 Online Learning Platform Interaction Behavior Analysis and Real-Time Feedback Algorithm with LSTM Temporal Modeling and Attention Mechanism Integration
Year: 2026
Type: conference-paper
Abstract: To address the feedback lag problem in K12 (Kindergarten to 12th grade) online learning caused by the difficulty of capturing dynamic behavioral sequences using traditional analysis methods, this paper proposes a real-time analysis algorithm that integrates LSTM (Long Short-Term Memory) temporal modeling and an attention mechanism. First, the platform interaction logs are encoded and sequence-normalized. Then, LSTM is used to capture the temporal evolution of learning states. Next, mechanisms are introduced to enhance focus on key behavioral components and improve the interpretability of the model. Finally, a real-time feedback strategy is generated based on dynamic probability thresholds and a rule engine. Experiments show that this method achieves a state recognition accuracy of 85.3%, a key behavior localization accuracy exceeding 83.5%, and an average system feedback latency of only 210 milliseconds. It effectively realizes a closed loop from behavior analysis to immediate intervention, providing an interpretable and low-latency solution for intelligent learning tutoring systems.
Author: Junxia Zheng
Institution: Shandong Judicial Police Vocational College
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
SDG: Quality education
Open Access status: hybrid
OpenAlex ID: W7166669578