Journal of Computer Applications ›› 2025, Vol. 45 ›› Issue (7): 2043-2055.DOI: 10.11772/j.issn.1001-9081.2024070970

• The 39th CCF National Conference of Computer Applications (CCF NCCA 2024) •     Next Articles

Review of interpretable deep knowledge tracing methods

Jinxian SUO, Liping ZHANG(), Sheng YAN, Dongqi WANG, Yawen ZHANG   

  1. College of Computer Science and Technology,Inner Mongolia Normal University,Hohhot Inner Mongolia 010022,China
  • Received:2024-07-03 Revised:2024-09-06 Accepted:2024-09-25 Online:2025-07-10 Published:2025-07-10
  • Contact: Liping ZHANG
  • About author:SUO Jinxian, born in 2001, M. S. candidate. His research interests include educational data mining,smart education.
    YAN Sheng, born in 1984,M. S., lecturer. His research interests include computer education applications.
    WANG Dongqi, born in 1997, M. S. candidate. His research interests include educational data mining,smart education.
    ZHANG Yawen, born in 1998, M. S. candidate. Her research interests include educational data mining,code analysis.
  • Supported by:
    National Natural Science Foundation of China(61462071);Inner Mongolia Natural Science Foundation(2023LHMS06009);Inner Mongolia Autonomous Region Education Science Research “14th Five Year Plan” Annual Project(2023NGHZXZH119)

可解释的深度知识追踪方法综述

索晋贤, 张丽萍(), 闫盛, 王东奇, 张雅雯   

  1. 内蒙古师范大学 计算机科学技术学院,呼和浩特 010022
  • 通讯作者: 张丽萍
  • 作者简介:索晋贤(2001—),男,山西大同人,硕士研究生,CCF学生会员,主要研究方向:教育数据挖掘、智慧教育
    闫盛(1984—),男,内蒙古包头市人,讲师,硕士,CCF会员,主要研究方向:计算机教育应用
    王东奇(1997—),男,内蒙古赤峰人,硕士研究生,CCF学生会员,主要研究方向:教育数据挖掘、智慧教育
    张雅雯(1998—),女,内蒙古阿拉善盟人,硕士研究生,CCF学生会员,主要研究方向:教育数据挖掘、代码分析。
  • 基金资助:
    国家自然科学基金资助项目(61462071);内蒙古自然科学基金资助项目(2023LHMS06009);内蒙古自然科学基金资助项目(2024MS06020);内蒙古自治区教育科学研究“十四五”规划2023年度课题(2023NGHZXZH119);内蒙古自治区教育科学研究“十四五”规划2023年度课题(NGJGH2023234)

Abstract:

Knowledge Tracing (KT) is a cognitive diagnostic method aimed at simulating learner's mastery level of learned knowledge by analyzing learner's historical question answering records, ultimately predicting learner's future question answering performance. Knowledge tracing techniques based on deep neural network models have become a hot research topic in knowledge tracing field due to their strong feature extraction capabilities and superior prediction performance. However, deep learning-based knowledge tracing models often lack good interpretability. Clear interpretability enable learners and teachers to fully understand the reasoning process and prediction results of knowledge tracing models, thus facilitating the formulation of learning plans tailored to the current knowledge state for future learning, and enhance the trust of learners and teachers in knowledge tracing models at the same time. Therefore, interpretable Deep Knowledge Tracing (DKT) methods were reviewed. Firstly, the development of knowledge tracing and the definition as well as necessity of interpretability were introduced. Secondly, improvement methods proposed for solving the lack of interpretability in DKT models were summarized and listed from the perspectives of feature extraction and internal model enhancement. Thirdly, the related publicly available datasets for researchers were introduced, and the influences of dataset features on interpretability were analyzed, discussing how to evaluate knowledge tracing models from both performance and interpretability perspectives, and sorting out the performance of DKT models on different datasets. Finally, some possible future research directions to address current issues in DKT models were proposed.

Key words: smart education, Deep Knowledge Tracing (DKT), explainability, knowledge tracing model, deep learning

摘要:

知识追踪(KT)是一种认知诊断方法,旨在通过学习者的历史答题记录,模拟学习者对于学习知识的掌握程度,最终预测学习者未来的答题情况。目前基于深度神经网络模型的知识追踪技术以强大的特征提取能力和优越的预测能力成为知识追踪领域研究的热点;然而,基于深度学习的知识追踪模型通常缺乏较好的可解释性。清晰的可解释性不仅可以让学习者和教师充分理解知识追踪模型的推理过程和预测结果,从而为下一步学习制定符合当前知识状态的学习计划,还能够提升学习者和教师对知识追踪模型的信任程度。因此,对可解释的深度知识追踪(DKT)方法进行综述。首先,介绍知识追踪的发展历程,并介绍可解释性的定义和必要性;其次,从特征提取和模型内部提升两方面,对解决DKT模型缺乏可解释性而提出的改进方法进行总结和梳理;再次,介绍现有的可供研究者使用的相关公开数据集,并分析数据集内的数据特征对可解释性的影响,从而探讨如何从模型性能和可解释性两个方面对知识追踪模型进行评价,并整理DKT模型在不同数据集上的性能表现;最后,对DKT模型目前存在的问题提出一些未来可能的研究方向。

关键词: 智慧教育, 深度知识追踪, 可解释性, 知识追踪模型, 深度学习

CLC Number: