知识追踪旨在评估学习者的学习状态,并根据先前的答题情况预测他们未来的答题表现.然而现有的知识追踪模型大多仅关注问题或技能间的关联,忽略了学生与问题间的结构关系.为此我们提出了基于学生问题关联的异构图知识追踪模型(StudentPro...知识追踪旨在评估学习者的学习状态,并根据先前的答题情况预测他们未来的答题表现.然而现有的知识追踪模型大多仅关注问题或技能间的关联,忽略了学生与问题间的结构关系.为此我们提出了基于学生问题关联的异构图知识追踪模型(StudentProblem association based heterogeneous graph Knowledge Tracing model,SPKT).该模型在知识追踪中融合了学生的学习能力和问题的重要性,并使用图注意力网络学习学生问题间的关联,获得学生、问题的嵌入表示并进行知识状态的预测.通过在真实公开数据集上的性能对比和模型消融实验,并可视化SPKT模型的知识追踪效果,证明了SPKT在预测性能上优于现有的知识追踪模型.展开更多
A new method for combining features via importance-inhibition analysis (IIA) is described to obtain more effective feature combination in learning question classification. Features are combined based on the inhibiti...A new method for combining features via importance-inhibition analysis (IIA) is described to obtain more effective feature combination in learning question classification. Features are combined based on the inhibition among features as well as the importance of individual features. Experimental results on the Chinese questions set show that, the IIA method shows a gradual increase in average and maximum accuracies at all feature combinations, and achieves great improvement over the importance analysis(IA) method on the whole. Moreover, the IIA method achieves the same highest accuracy as the one by the exhaustive method, and further improves the performance of question classification.展开更多
文摘知识追踪旨在评估学习者的学习状态,并根据先前的答题情况预测他们未来的答题表现.然而现有的知识追踪模型大多仅关注问题或技能间的关联,忽略了学生与问题间的结构关系.为此我们提出了基于学生问题关联的异构图知识追踪模型(StudentProblem association based heterogeneous graph Knowledge Tracing model,SPKT).该模型在知识追踪中融合了学生的学习能力和问题的重要性,并使用图注意力网络学习学生问题间的关联,获得学生、问题的嵌入表示并进行知识状态的预测.通过在真实公开数据集上的性能对比和模型消融实验,并可视化SPKT模型的知识追踪效果,证明了SPKT在预测性能上优于现有的知识追踪模型.
基金The National Natural Science Foundation of China(No.61003112,61170181)the Open Research Fund of State Key Laboratory for Novel Softw are Technology of China(No.KFKT2010B02)the Key Project of Natural Science Research for Anhui Colleges of China(No.KJ2011A048)
文摘A new method for combining features via importance-inhibition analysis (IIA) is described to obtain more effective feature combination in learning question classification. Features are combined based on the inhibition among features as well as the importance of individual features. Experimental results on the Chinese questions set show that, the IIA method shows a gradual increase in average and maximum accuracies at all feature combinations, and achieves great improvement over the importance analysis(IA) method on the whole. Moreover, the IIA method achieves the same highest accuracy as the one by the exhaustive method, and further improves the performance of question classification.