Research on Optimization of Blended Learning Paths Based on TPACK Framework and Deep Reinforcement Learning
Year: 2026
Type: conference-paper
Abstract: The rapid development of digital education and smart teaching environments is driving the evolution of blended learning towards intelligence and personalization. However, traditional teaching paths still have shortcomings in resource matching, dynamic feedback, and learning adaptation. This paper constructs a blended learning path optimization model based on the TPACK framework and deep reinforcement learning. By integrating technological knowledge, pedagogical knowledge, and subject knowledge, it dynamically models student learning behavior, classroom interaction, and resource allocation, and uses reinforcement learning algorithms to achieve adaptive optimization of teaching strategies. Experimental results show that the proposed model outperforms traditional teaching models in terms of average score, classroom participation, and task completion rate. Specifically, the average student score reached 88.7 points, classroom participation reached 89.5%, and the task completion rate reached 91.2%, representing improvements of 14.1 points, 21.2%, and 19.7% respectively compared to traditional blended learning. Furthermore, the model convergence rounds decreased to 185, and the path optimization accuracy reached 92.8%, indicating that the integration of TPACK and deep reinforcement learning can effectively improve the efficiency of teaching path optimization and teaching quality, providing new research ideas for personalized teaching decisions in the context of smart education.
Author: Ying Han
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
SDG: Quality education
Open Access status: hybrid
OpenAlex ID: W7201875672