基于改进Apriori算法的高校教育满意度关联规则挖掘  

Association Rule Mining of Higher Education Satisfaction Based on Improved Apriori Algorithm

作　　者：陈云超 谢加良[1,2] 林玲 刘小辉 CHEN Yunchao;XIE Jialiang;LIN Ling;LIU Xiaohui(School of Science,Jimei University,Xiamen 361021,China;Digital Fujian Big Data Modeling and Intelligent Computing Institute,Xiamen 361021,China;Institute of Educational Research,Xiamen University,Xiamen 361005,China)

机构地区：[1]集美大学理学院,福建厦门361021 [2]数字福建大数据建模与智能计算研究所,福建厦门361021 [3]厦门大学教育研究院,福建厦门361005

出　　处：《集美大学学报（自然科学版）》2024年第4期377-384,共8页Journal of Jimei University：Natural Science

基　　金：全国教育科学规划2021年度国家一般课题“高校在线教育高质量发展模式研究”(BIA210171)。

摘　　要：针对经典关联规则Apriori算法在大数据集情境下易产生冗余和误导性的关联规则,以及难以确认关键性关联规则等问题,提出支持度—置信度—权重检验系数框架与后项约束的改进Apriori算法。首先,定义相关性系数、提升系数、错误系数并进行证明分析,进而构建权重检验系数;其次,运用主成分分析法,提取指标中的高权重影响因素作为后项,通过后项约束过滤冗余关联信息,从而筛选出更为准确的关键性关联规则。将改进的Apriori算法应用于高校教育满意度调查数据的关联规则挖掘并进行分析对比,实验结果验证了该算法的合理性和有效性。Due to the problem that the classical association rule Apriori algorithm is prone to produce redundant and misleading association rules in the context of large data sets,and it is difficult to identify key association rules,this paper proposes an improved Apriori algorithm based on the support-confidence weight-test coefficient framework and the post-term constraint.Firstly,the correlation coefficient,lifting coefficient and error coefficient were defined and proved,and then the weight test coefficient was constructed.Secondly,the principal component analysis method is used to extract the influential factors with high weight in the index as the latter term,and the redundant association information is filtered through the latter term constraints,so as to screen out more accurate key association rules.The improved Apriori algorithm is applied to mining association rules of the survey data of higher education satisfaction,and the experimental results verify the rationality and effectiveness of the algorithm.

关 键 词：高校教育满意度 数据挖掘 关联规则 APRIORI算法 

分 类 号：G434[文化科学—教育学] TP311[文化科学—教育技术学]