Research on Blended Learning Behavior Based on Attention Mechanism

The research on blended learning mainly focuses on hybrid teaching mode, flipped classroom and MOOC etc. Most of researches address online learning behaviors, instead of learning behaviors in hybrid learning environment. Research on blended learning behavior seeks to uncover some effective ways to harness the benefits of both face-to-face and online learning. This study collects the data from the teaching practice in a blended learning environment. A Spearman correlation analysis model is proposed to explore the impact of students' learning behavior on performance. Then, based on students' deep learning behavioral characteristics and attention mechanisms, a learning performance prediction model is constructed and applied to the experimental verification of the learning behavior data generated by the application of blended teaching mode in curriculum. Experimental results indicates that the proposed model can accurately predict students' performance based on their blended learning behaviors, thus providing instructors with countermeasures to improve students' learning behavior and further enhance the blended learning effect.