SBPM Model for Analyzing Students’ Learning Behavior Based on Fine Grained Emotion Analysis and Emotion Assessment

Different students have different learning behaviors, and their attitudes towards learning are directly reflected in their learning behavior. To improve students' self-learning ability and enhance the teaching level of universities, this study constructs an emotional change trend evaluation model based on fine-grained emotion analysis technology during the learning process. Based on the output results of the model, predictive analysis is conducted on students' learning behavior. The results showed that the designed emotional evaluation model could achieve an accuracy of over 80% in analyzing the trend of students' emotional changes, and the calculation time was only about 15 seconds. The proposed student learning behavior prediction model could reduce the average percentage error to 0.15% when predicting and analyzing students’ learning behavior. The proposed student learning behavior prediction model consistently maintained an F1 score above 0.95 and an accuracy rate of over 97% in predicting students’ learning behavior. The research model can accurately analyze the emotional changes of students during the learning process and predict and analyze their learning behavior. Universities can correct students' learning behavior based on the output results of the model, effectively improving students' learning efficiency and enhancing the teaching level of universities.