面向应用题的问题情境理解能力追踪模型研究  

Research on the tracking model of problem situational understanding ability for word problems

作　　者：余小鹏[1] 王振佩 殷浩 YU Xiaopeng;WANG Zhenpei;YIN Hao(School of Management,Wuhan Institute of Technology,Wuhan,China 430205;Wuhan Jingkai Foreign Language School,Wuhan,China 430010)

机构地区：[1]武汉工程大学管理学院,湖北武汉430205 [2]武汉经开外国语学校,湖北武汉430010

出　　处：《长江信息通信》2022年第9期121-123,共3页Changjiang Information & Communications

基　　金：2019年度教育部人文社会科学研究规划基金项目“基于问题情境仿真的数学应用题表征辅导系统研究”(编号:19YJA880077)。

摘　　要：问题情境的理解是应用题表征的关键,运用贝叶斯知识追踪模型BKT,对被试的问题情境理解能力进行建模,有利于对被试的个性化辅导。问题情境理解属于被试的心理活动,不易显式展示,使得用于BKT分析所需的数据难以收集。文章提出问题情境理解能力追踪模型PS-BKT,通过被试对问题情境的重构,显示展示其对问题情境的理解,并收集相关数据。实验结果表明:PS-BKT模型能较好收集数据,较快估算模型参数。The understanding of problem situation is the key to the representation of word problem. The BKT is used to model the ability of problem situation understanding of the subjects, which is beneficial to the personalized guidance of the subjects. Problem situation understanding belongs to the mental activity of the subject and is not easy to be displayed explicitly, which makes it difficult to collect the data required for BKT analysis. PS-BKT, a tracking model of problem situation comprehension ability, is proposed. Through the reconstruction of problem situation, participants’ understanding of problem situation is displayed and relevant data are collected. Experimental results show that PS-BKT model can collect data well and estimate model parameters quickly.

关 键 词：知识追踪 应用题 智能导学系统 贝叶斯 

分 类 号：C931.6[经济管理—管理学]