多知识点融合嵌入的深度知识追踪模型
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

TP311

基金项目:

国家自然科学基金重点项目(62137001)


A Deep Knowledge Tracing Model with Multi-Concept Embedding
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    知识追踪任务是根据学生历史答题记录追踪学生知识状态的变化,预测学生未来的答题情况.近年来,基于注意力机制的知识追踪模型在灵活性和预测性能上都明显优于传统知识追踪模型.但是现有深度模型大多只考虑了单一知识点题目的情况,无法直接处理多知识点题目,而智能教育系统中存在着大量的多知识点题目.此外,如何提高可解释性是深度知识追踪模型的关键挑战之一.为了解决这些问题,论文提出了一种多知识点融合嵌入的深度知识追踪模型.论文模型考虑了涉及多知识点的题目中知识点之间的关系,提出了两种新颖的多知识点嵌入方式,并且结合教育心理学模型和遗忘因素提升预测性能和可解释性.实验表明了论文模型在大规模真实数据集上预测性能上优于现有模型,并验证了论文模型各个模块的有效性.

    Abstract:

    The task of knowledge tracing is to trace the changes of students' knowledge state and predict their future performance in learning according to their historical learning records. In recent years, knowledge tracing models based on attention mechanism are obviously superior to traditional knowledge tracing models in both flexibility and prediction performance. However, only the situation of single concept is considered by most of the existing deep models which cannot directly deal with multi-concept exercises, while there are a large number of multi-concept exercises in the intelligent education systems. In addition, how to improve interpretability is one of the key challenges of the deep knowledge tracing models. To solve these problems, this paper proposes a deep knowledge tracing model with multi-concept enhanced exercise embedding, which considers the relationship between concepts in exercises and proposes two novel multi-concept embedding methods and combines educational psychology models with forgetting factors to improve prediction performance and interpretability. Experiments have shown that the model proposed by this paper outperforms existing models in predictive performance on large-scale real datasets, and the effectiveness of each module is verified.

    参考文献
    相似文献
    引证文献
引用本文

琚生根,康睿,赵容梅,孙界平.多知识点融合嵌入的深度知识追踪模型.软件学报,,():0

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2022-02-21
  • 最后修改日期:2022-05-25
  • 录用日期:
  • 在线发布日期: 2022-12-30
  • 出版日期:
您是第位访问者
版权所有:中国科学院软件研究所 京ICP备05046678号-3
地址:北京市海淀区中关村南四街4号,邮政编码:100190
电话:010-62562563 传真:010-62562533 Email:jos@iscas.ac.cn
技术支持:北京勤云科技发展有限公司

京公网安备 11040202500063号