基于试题关系树的自适应诊断测试模型

Adaptive learning diagnosis model based on question relationship tree

针对实际测试过程中试题可能包含多个知识点组合的问题,提出一种基于试题关系树的自适应诊断测试模型的解决方案。根据知识点组合的包含关系,分析得出试题之间存在的3种前提关系,在此基础上提出以与子结点和或子结点将试题库进行关联,形成试题关系树。重点分析了基于试题库构建试题关系树的过程,给出了基于试题关系树的测试策略与诊断过程,并最终形成诊断决策树。实验结果表明,该模型可有效地减小平均测试长度,提高诊断效率。

Considering the fact that a question could have more than one compounds of knowledge point in actual tests,an adaptive learning diagnosis mode based on the question relationship tree was proposed.According to the including relationships of questions,three kinds of premise relationship were analyzed,based on which questions could be connected by and-sub-node and or-sub-node,then the relationship tree was finally built.The process how to build a relationship tree from the question bank was analyzed.Finally,the testing strategy and the process of diagnosis were introduced and the diagnosis decision tree was built.Simulation experiment showed that the model could efficiently reduce the average length of testing and has good effect on the efficiency of testing.