基于教育数据挖掘的学业表现影响因素研究

Research on Academic Performance Base on Educational Data Mining

本研究运用数据挖掘技术对高校学生学业表现和影响因素进行分析,目的在于提高学生学业表现和改善教学模式。首先采集学生的人口统计、学习环境、个人特征和学习投入四个方面的信息,构建综合教育数据框架,然后通过使用SPSS Modeler软件分别建立决策树、贝叶斯网络、神经网络和支持向量机四种预测模型来探讨影响学业表现的因素,最后以湖北某高校为例进行实证研究,通过比较模型的准确率、召回率和F1值对模型有效性进行评估。实验结果表明,习题完成情况、每周学习时长和性别是影响学生学业表现最重要的三大因素。

This study uses data mining technology to analyze college students'academic performance and influencing factors,in order to improve students'academic performance and improve teaching mode.First,collect the information of students'demographics,learning environment,personal characteristics and learning investment,and build a comprehensive education data framework.Then,four prediction models of decision tree,Bayesian network,neural network and support vector machine are established by using SPSS Modeler software to explore the factors affecting academic performance.Finally,a university in Hubei is taken as an example to evaluate the effectiveness of the model by comparing the accuracy,recall and F1 value of the model.The experimental results show that the completion of exercises,the length of study per week and gender are the three most important factors affecting students'academic performance.