qDKT：以数据库为中心的深度知识追踪
qDKT: Question-centric Deep Knowledge Tracing
摘要：
知识追踪（KT）模型，例如，深度知识追踪（DKT）模型，通过检查学习者在与这些技能相关的问题上的表现来跟踪个人学习者随着时间的推移对技能的习得。在大多数现有的KT模型中的实际限制是，嵌套在特定技能下的所有问题都被视为学习者能力的等效观察，这在现实世界的教育场景中是不准确的假设。为了克服这个限制，我们引入了qDKT，这是DKT的一个变体，它可以随着时间的推移模拟每个学习者在各个问题上的成功概率。首先，qDKT结合了图拉普拉斯正则化来平滑每个技能下的预测，这在数据集中的问题数量很大时特别有用。其次，qDKT使用了一个受fastText算法启发的初始化方案，该算法在各种语言建模任务中取得了成功。我们在几个真实世界数据集上的实验表明，qDKT在预测学习者结果方面达到了最先进的性能。正因为如此，qDKT可以作为一个简单的，但难以击败的，新的以问题为中心的KT模型的基线。
英文摘要：
Knowledge tracing (KT) models, e.g., the deep knowledge tracing (DKT) model, track an individual learner's acquisition of skills over time by examining the learner's performance on questions related to those skills. A practical limitation in most existing KT models is that all questions nested under a particular skill are treated as equivalent observations of a learner's ability, which is an inaccurate assumption in real-world educational scenarios. To overcome this limitation we introduce qDKT, a variant of DKT that models every learner's success probability on individual questions over time. First, qDKT incorporates graph Laplacian regularization to smooth predictions under each skill, which is particularly useful when the number of questions in the dataset is big. Second, qDKT uses an initialization scheme inspired by the fastText algorithm, which has found success in a variety of language modeling tasks. Our experiments on several real-world datasets show that qDKT achieves state-of-art performance on predicting learner outcomes. Because of this, qDKT can serve as a simple, yet tough-to-beat, baseline for new question-centric KT models.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
