基于数据挖掘技术的高校教学质量监控研究  

Research on Teaching Quality Monitoring in Universities Based on Data Mining Technology

作　　者：郭欣[1] 章鸣嬛[1] 吴良 陈瑛[1] GUO Xin;ZHANG Minghuan;WU Liang;CHEN Ying(Research Center of Big Data Analysis and Process,School of Information Science and Technology,Shanghai Sanda University,Shanghai 201209;Teaching Quality Management Department,Shanghai Sanda University,Shanghai 201209)

机构地区：[1]上海杉达学院信息科学与技术学院大数据分析与处理研究中心,上海201209 [2]上海杉达学院教学质量管理处,上海201209

出　　处：《微型电脑应用》2020年第1期7-10,21,共5页Microcomputer Applications

基　　金：2018年教育部高教司协同育人项目(201802001049);2018年教育部高教司协同育人项目(201802111036)

摘　　要：教学质量监控是学校进行教学质量管理的重要手段,对提高教学质量具有重要意义。随着校园信息化的发展,高校教务系统里积存了大量有价值的数据,但却没有被很好的挖掘利用。鉴此提出将数据挖掘技术应用到高校教学质量监控中,使用因子分析法对计算机专业的学生成绩进行综合评价分析,并提出一种决策树改良方法对学生成绩进行预测,最后针对文中所提出的方法与其他方法进行了对比分析。结果表明,因子分析法相对于平均分排名法,蕴含了更多的信息量可以提供更加全面的评价,另外其突出重要因素,可以对各方面情况进行合理量化;决策树改良方法相对于其他成绩预测方法稳定性好、准确性高,且过拟合情况少。经验证,这个方法可以在高校教学质量监控中发挥一定作用。Monitoring teaching quality is an important means in teaching quality management in schools,and it is of great significance to improve the teaching quality.With the development of campus informatization,a lot of valuable data have been accumulated in the educational administration system of universities,but they have not been well mined and utilized.So this paper puts forward the application of data mining technology to the monitoring of teaching quality in universities.The factor analysis method is used to make a comprehensive evaluation and analysis of the students’achievements in computer major,and a decision tree improvement method is proposed to predict students’achievements.Finally,the methods proposed in this paper are compared with other methods in detail.The results show that factor analysis method contains more information than traditional methods and can provide more comprehensive evaluation.In addition,it highlights important factors and can quantify all aspects of the situation reasonably.Compared with the other methods,the improved method of decision tree has better stability,higher accuracy and less over-fitting.It has been proved that the method proposed in this paper can play a certain role in the monitoring of teaching quality in universities.

关 键 词：数据挖掘技术 教学质量监控 因子分析 决策树 聚类 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]