Abstract
Abstract
Background
The rapid advancement of educational informatization has positioned online learning as a mainstream educational modality. Nevertheless, persistent challenges — particularly low course completion rates and high learner attrition — continue to constrain its effectiveness.
Objective
This study systematically examines online learning behaviors across three core dimensions — learning duration, course completion rate, and grade distribution — with the aim of identifying actionable strategies to enhance platform performance.
Methods
A mixed-methods design was adopted, integrating literature review, empirical analysis, and case studies. A theoretical framework was constructed, and a behavioral feature set encompassing learning duration, completion rates, grade distribution, and interaction metrics was established. On this basis, predictive models were developed and benchmarked using classification algorithms, ensemble learning, deep learning, and clustering techniques.
Results
Findings indicate a statistically significant positive correlation among learning duration, course completion rates, and grade distribution, confirming their roles as key indicators of online learning effectiveness. From a methodological standpoint, behavior analysis has progressed toward advanced deep learning architectures, with reported prediction accuracies surpassing 95% on benchmark datasets. Distinct behavioral patterns were identified among learners — categorized as high-engagement, low-engagement, skip-pattern, and strategic types. Additionally, instructional design was found to exert a notable influence on behavioral outcomes.
Conclusion
Drawing on these findings, optimization strategies are proposed at three levels: (1) Platform level — strengthen data analytics infrastructure to support early-warning systems and personalized recommendation mechanisms; (2) Instructional design level — prioritize scenario-based learning, resource diversification, interaction enhancement, and feedback optimization; (3) Management level — reinforce behavioral monitoring, implement early-intervention protocols, facilitate instructor training, and establish incentive mechanisms. This study provides an evidence-based foundation for personalized recommendation and curriculum design, offering practical implications for advancing the quality of online education.