Analysis of Student Learning Willingness

Citation Author(s):
Submitted by:
Hui Mao
Last updated:
Tue, 03/19/2024 - 21:55
0 ratings - Please login to submit your rating.


Student learning willingness is the decisive factor for achieving the final learning outcomes in curriculum teaching. On the other hand, the final learning outcomes achieved by students in the curriculum are a true reflection of student learning willingness. This paper selects 6 types of theoretical teaching method data and 4 types of student engagement behavior data used in the teaching process of the "Computer Systems" course in the Software Engineering major of Information Engineering School in the academic years 2021, 2022, and 2023 as the basic data. The purpose of this research is to utilize daily authentic data recorded by classroom “micro-assistant” platform, introduce interpretable machine learning methods into the field of education, and analyze the various factors affecting students' willingness to learn in theoretical course teaching, exemplified by the "Computer Systems" course, and explain their contributions.


This paper uses the real data recorded by the micro-assisted teaching platform in the daily teaching process of the course "Computer Systems" for students majoring in Software Engineering at Information Engineering School in 2021, 2022, and 2023. This can provide better guidance and reference for future actual teaching. Based on existing research, this paper selects 10 basic indicators listed in table I: score of class attendance rates (attend_score), average score of post-class assignments (hw_score), average score of group activities or discussion (discus_score), average score of classroom question answering (ans_score) as four student participation behavioral data indicators, and whether smart classrooms are being used for teaching (smart_cr), the frequency of using gaming teaching methods such as storytelling and video viewing (s&v_num), the frequency of using group activities or discussion (discus_num), the number of post-class assignments (hw_num), the frequency of classroom questions (ans_num) and the percentage of actual teacher explanations (explan_t) as the six indicators of multimodal teaching methods. Using these 10 indicators as input features, and classifying them into low / medium / high three categories of learning willing states (0, 1, 2) as output labels, to establish a predictive model.