基于演化聚类的学习者知识跟踪模型研究

Student Knowledge Tracing Model Based on Evolutionary Clustering

传统的知识跟踪模型存在个性化能力不足和预测精度差等问题,难以同时促进学习者的进步和教学系统的改进。为此,提出一种基于演化聚类的学习者知识跟踪模型。该模型将学习者在智能教学系统中的交互数据按照时间推移进行动态聚类,得到不同知识水平的学生群体,并传入贝叶斯知识跟踪模型得到预测结果。模型充分考虑学习者的个体差异以及学习者知识水平变化的时间平滑特性,能有效缓解异常数据的干扰,能更好地服务于学习者和教学系统。

The traditional knowledge tracing model has some limitations,such as lack of personalized ability and poor prediction accuracy,which is difficult to promote the progress of learners and the improvement of tutoring system at the same time.Therefore,a student knowledge tracing model based on evolutionary clustering is proposed.In this model,the interactive data of students in the intelligent tutoring system are dynamically clustered over time to obtain the student groups with different knowledge levels,and transfers them into the bayesian knowledge tracing model to obtain prediction results.The model fully considers the individual differences of the learners and the temporal coherence of the knowledge levels of the students.This model can effectively alleviate the interference of abnormal data,and can better serve the learners and teaching systems.