COLNet: A Chinese Online Learning Analysis Model

With the rise of China's international status and the growing global enthusiasm for learning Chinese, online education, as a flexible and convenient teaching model, has become an indispensable part of modern education systems, especially in the field of Chinese language teaching. However, learner engagement in online learning environments is influenced by more complex factors such as individual characteristics, teachers, peers, and environmental conditions. This study aims to explore the online learning engagement of Chinese learners and its influencing factors through deep learning models, with the goal of improving teaching quality and learning outcomes. We first conducted a survey to collect learners' personal information and based on this, built a deep learning model called COLNet to predict the online learning engagement of Chinese learners. Our dataset includes behavioral data from over 500,000 active users across 15 major Chinese language learning platforms from 2018 to 2023, with differential privacy processing to ensure compliance. The experimental results show that COLNet achieved state-of-the-art performance in both ten-class and five-class tasks, with significant improvements over other mainstream models. Additionally, ablation experiments validated the importance of each component for the overall model performance. This study not only provides new insights for the design of personalized education programs but also emphasizes the critical role of strategies such as optimizing online learning platform design and enhancing teachers' online teaching capabilities in improving the online learning experience for Chinese learners. Ultimately, we hope these findings will help Chinese language learners better adapt to the online learning environment, improving learning efficiency and outcomes.