基于数据挖掘的虚拟仿真实验教学量化分析  

Quantitative Analysis of Virtually Experimental Teaching Using Data Mining

作　　者：张浩[1] 雷洪[1] 龚成斌[1] 彭敬东[1] 何显达 马学兵[1] 

机构地区：[1]西南大学化学化工学院,重庆400715

出　　处：《实验室研究与探索》2017年第9期129-131,144,共4页Research and Exploration In Laboratory

基　　金：西南大学博士基金项目(SWU114021);西南大学基本科研业务费(XDJK2015C101)

摘　　要：作为实际实验教学的重要补充,基于互联网平台的虚拟仿真实验教学实践过程中生成了大量的学生学习行为数据。为了研究虚拟仿真实验教学与实际实验教学成绩的关系,识别虚拟仿真实验教学中的新行为模式,利用数据挖掘技术对化学化工虚拟仿真实验教学过程数据进行了分析。通过研究虚拟-现实实验成绩之间的相关性、因果性以及学习时间对成绩的影响,发现理论学习、实际操作成绩与偏重于流程训练的简单虚拟仿真实验项目的相关性较小,与机理型复杂虚拟仿真实验项目相关性较大;虚拟仿真实验的开展对实际实验均有不同程度的促进作用。As an important supplement to real experimental teaching, there are a great amount of data about learning process during the execution of virtually experimental teaching based on the Internet platforms. To study relations between virtual experiment scores and real experiment scores and recognize new patterns of learning behaviors in virtual experiment teaching process, data mining techniques are employed to analyze data of virtual chemical and chemical engineering experiment teaching processes. The results of relations and causalities between virtual experiment scores and real experiment scores and impact of learning time on scores show that relations between theoretical learning, real experiment scores and simple virtual experiment programs about process training are much lower than that with complicated virtual experiment programs based on mechanism. Executions of virtual experiments can improve their performances on real experiment in different degrees.

关 键 词：虚拟仿真 数据挖掘 量化分析 实验教学 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] G642.1[自动化与计算机技术—计算机科学与技术]