学习者在线动态学习成绩的Logistic曲线增长规律及其聚类研究  

To Cluster the Learners′Online Dynamic Learning Performance Based on the Logistic Curve Growth Model

作　　者：熊思灿 农莹[2] XIONG Sican;NONG Ying(School of Economics and Management,East China University of Technology,Nanchang 330013,China;School of Education,Central China Normal University,Wuhan 430079,China)

机构地区：[1]东华理工大学经济与管理学院,江西南昌330013 [2]华中师范大学教育学院,湖北武汉430079

出　　处：《东华理工大学学报（社会科学版）》2024年第1期89-94,共6页Journal of East China University of Technology(Social Science)

基　　金：东华理工大学校级教育科学规划课题“数据驱动下的智能化在线教学全过程管理与评价”(22XYB22)。

摘　　要：基于信息化教学中的动态学习数据,采用Logistic曲线拟合学习者的综合在线学习成绩随学习时长的增长规律,可将无穷维函数空间转化为有限维的参数空间,并借助于经典的K-Means聚类算法实现学习者的综合在线学习成绩增长规律的聚类。研究结果表明:学习者可被分为主流学习者和非主流学习者两大类别;主流学习者占全部学习者的96.92%,且可被进一步细分为高效学习者、中效学习者、低效学习者,占全部学习者的比例分别为16.41%、61.02%和19.49%;非主流学习者可被分为长期缓增型学习者和短期陡增型学习者,占全部学习者的比例均为1.54%。此外,不同类别学习者之间的期末考试平均成绩存在显著差异,主流学习者中高效学习者、中效学习者和低效学习者的期末考试平均成绩逐渐递减,但均高于非主流学习者的期末考试平均成绩;在非主流学习者中,短期陡增型学习者的期末考试平均成绩高于长期缓增型学习者的期末考试平均成绩。Based on the dynamic learning data in information-based teaching,the Logistic curve is used to fit the growth law of learners′comprehensive online learning performance with the learning time.By doing that,the infinite-dimensional function space is transformed into a finite-dimensional parameter space,and the clustering of the growth law of learners′comprehensive online learning performance is realized by using the classical K-Means clustering algorithm.The results show that learners can be divided into mainstream learners and non-mainstream learners.Among them,mainstream learners account for 96.92%of all learners,that is,the vast majority,and can be further subdivided into"high-efficient learners""moderate-efficient learners"and"low-efficient learners",which account for about 16.41%,61.02%and 19.49%of all learners,respectively.Non-mainstream learners can be further divided into"long-term slow increase type"learners and"short-term rapid increase type"learners,and both account for 1.54%of all learners.In addition,there is a very significant difference in the average final exam scores between different categories of learners,and the average final exam scores of the"high""medium"and"low"efficient mainstream learners gradually decrease,but they are all higher than those of non-mainstream learners.Among non-mainstream learners,the average final exam scores of learners with"short-term rapid increase type"is higher than those of"long-term slow increase type"learners.

关 键 词：综合在线学习成绩 Logistic曲线 两步串联曲线聚类法 肘法 

分 类 号：G420[文化科学—课程与教学论]