The goal of knowledge tracing is to track the state of a student's knowledge as it evolves over time. This plays a fundamental role in understanding the learning process and is a key task in the development of an intelligent tutoring system. In this paper we propose a novel approach to knowledge tracing that combines techniques from matrix factorization with recent progress in recurrent neural networks (RNNs) to effectively track the state of a student's knowledge. The proposed \emph{DynEmb} framework enables the tracking of student knowledge even without the concept/skill tag information that other knowledge tracing models require while simultaneously achieving superior performance. We provide experimental evaluations demonstrating that DynEmb achieves improved performance compared to baselines and illustrating the robustness and effectiveness of the proposed framework. We also evaluate our approach using several real-world datasets showing that the proposed model outperforms the previous state-of-the-art. These results suggest that combining embedding models with sequential models such as RNNs is a promising new direction for knowledge tracing.
AlignKT: Explicitly Modeling Knowledge State for Knowledge Tracing with Ideal State Alignment
School of Computer Science South China Normal University Guangzhou, China（计算机科学学院 华南师范大学 广州, 中国）；School of Artificial Intelligence South China Normal University Foshan, China（人工智能学院 华南师范大学 广州, 中国）
Enhancing Deep Knowledge Tracing with Auxiliary Tasks
Jinan University（暨南大学）；Guangdong Institute of Smart Education（广东省智慧教育研究院）；TAL Education Group（好未来教育集团）；Think Academy International Education（学而思国际教育）
知识追踪（KT）是基于学生与智能教学系统的历史交互来预测学生未来表现的问题。最近的研究已经应用了多种类型的深度神经网络来解决KT问题。然而，现实世界的教育数据中有两个重要因素没有得到很好的体现。首先，大多数现有的作品增加输入表示的问题和知识成分的共现矩阵\footnote{\label{ft:kc}A KC is a generalization of everyday terms like concept, principle, fact, or skill.}（KCs），但没有明确地将这种内在的关系到最终的响应预测任务。第二，学生的个性化历史表现没有得到很好的捕捉。在本文中，我们提出了\emph{AT-DKT}来提高原始深度知识追踪模型的预测性能，并使用两个辅助学习任务，即，\emph{question tagging (QT) prediction task}和\emph{individualized prior knowledge (IK) prediction task}。具体来说，QT任务通过预测问题是否包含特定的KC来帮助学习更好的问题表征。IK任务通过逐步预测隐藏在学生历史学习交互中的学生水平先验知识来捕获学生的全局历史表现。我们在三个真实世界的教育数据集上进行了全面的实验，并将所提出的方法与深度序列KT模型和非序列模型进行了比较。实验结果表明，\emph{AT-D…
TutorLLM: Customizing Learning Recommendations with Knowledge Tracing and Retrieval-Augmented Generation
University of Southampton（索姆塞特大学）；Durham University（杜 ham 大学）；Japan Advanced Institute of Science and Technology（日本先进科学和技术研究院）；Newcastle University（新卡斯尔大学）
