| With the rapid development of campus informatization,campus card is widely used in all areas of student life, which not only made great student's campus life, but also have accumulated vast amounts of information, resulting in educational management, teaching research and other fields are facing big data problems.the most urgent task for scholars is that to explore the hidden valuable information in the data and model of these daily behavior data to predict the academic performance of students, which realize the exceptional conditions of early warning of the students, and make the appropriate intervention to make useful teaching decisions Information, and applied to teaching practice.In this paper, we use data mining to provide better information and guidance for college students' behavior and academic achievement. First of all, the paper analyzes the influence of the behavior analysis and behavior habit on the academic achievement in the educational theory. Secondly, the paper tap behavior data ,including user learning behavior, schedule and other characteristics from student daily .on the basis of data analysis, the characteristics of students' campus behavior will be revealed by data and analyze of students' behavior.then we use association analysis to excavate campus behavior and academic performance in the reading professional books, learning time arrangement and regular life habits,the results reflect the regular life habits are strongly associated with leaming.Finally,The student campus behavior data is modeled By the KNN algorithm ?SVM and Naive Bayesian algorithm comparison, the final selection of Naive Bayesian algorithm to predict the academic performance of students, the correct rate is 84.35%,We analyze the classification results from both global and local to assist information teaching,which can early warning abnormal situation for students, the results will play a role in the cultivation and management of students. |
PDF Full Text Request