Study on College Students’ Behavior Analysis System Based on Feature Extraction Algorithm Under Big Data Background

In order to improve the strategy formulation effect of intelligent teaching in colleges and universities, this paper extracts students’ learning characteristics under the background of big data, and analyzes students’ behaviors. Moreover, relying on the smart campus environment, this paper effectively, truly and comprehensively collects the space-time fine-grained daily behavior data generated by students on campus through intelligent sensing devices such as campus one-card and campus WiFi, and quantifies the daily behaviors acquired by students and the stability of their behavior environment. In addition, this paper collects the mental wandering frequency and content characteristics of students with different anxiety states in experimental situations by thinking sampling method, and verifies the performance of the system model proposed in this paper by combining with the solo experiment. The experimental results show that the model of studying students’ learning behavior proposed in this paper has more accurate recognition and analysis performance.