摘要
机器学习可帮助教师挖掘试卷中知识点之间的关系,检测学生知识点掌握情况,为评价反馈提供支持。使用深度学习Apriori算法,通过模拟研究和实证研究,针对具有不同知识点结构的试卷进行挖掘分析。模拟研究发现:Apriori算法能针对知识点属性复杂的试题进行知识点间关联规则的挖掘,准确率较高;随着样本量的增加,挖掘的准确率增加。实证研究发现:Apriori算法可对中小学的语文、数学、小学科学、中学物理等学科的试卷进行知识点间关联规则挖掘,但学科间挖掘结果有差异。经过粒度优化、删除基础知识点后,Apriori算法可较好挖掘中小学语文、数学剩余知识点间的关联关系,但对跨学科的知识点关系挖掘有待提升。
In recent years,some researchers have used machine learning to mine the relationship among the items knowledge and provide support for teachers'feedback.This study is based on Apriori algorithm,through simulation and empirical research,mining and analyzing test papers with different knowledge point structures.The simulation study shows that Apriori algorithm can mine the association rules among knowledge points for the test questions with complex attributes of knowledge points,and has a high accuracy.With the increase of samples,the accuracy of Apriori algorithm in mining test papers with complex knowledge points increases.The empirical study finds that Apriori algorithm can mine the association rules among knowledge points in the test papers of Chinese,mathematics,primary school science,middle school physics in primary and secondary schools.And the mining results of association rules among disciplines are different.After optimizing the granularity and deleting the basic knowledge points,Apriori algorithm can mine the association relationship between Chinese and mathematics residual knowledge points in primary and secondary schools,and the mining of interdisciplinary knowledge points in primary schools needs to be improved.
作者
赵宁宁
叶楠
陈小涵
王迪
温红博
Zhao Ningning;Ye Nan;Chen Xiaohan;Wang Di;Wen Hongbo(School of Chinese Language and Literature of Beijing Normal University,Beijing,100875;Shenzhen Longgang Tongxin Experimental School,Shenzhen,Guangdong,518116;China Basic Education Quality Monitoring Collaborative Innovation Center,Beijing Normal University,Beijing,100875;Chaoyang District Educatioanl Science Academy,Beijing,100028)
出处
《考试研究》
2024年第5期9-23,共15页
Examinations Research
基金
北京市教育科学规划“十三五”规划2020年度青年专项课题“STEAM教育的课程范式及其成效研究”(课题编号:CDCA2020099)的项目成果之一。
关键词
知识关联规则
APRIORI算法
学科
跨学科
不同知识点结构
Knowledge Association Rules
Apriori Algorithm
Disciplin
Interdisciplin
Differences in Knowledge Point Structure
