基于学科知识图谱的智能化认知诊断评估方法

An Intelligent Cognitive Diagnosis Assessment Method Based on Subject Knowledge Graph

认知诊断评估是新时代教育评价改革的重要着力点,而现有认知诊断评估方法存在认知模型构建效率低、可解释性差的问题,同时认知测量模型也因缺乏语义化的认知模型而导致精度不高。对此,文章提出基于学科知识图谱的智能化认知诊断评估(ICDA-SKG)方法,内容包含学科知识图谱和认知测量模型两个部分,其实现涉及学科知识点抽取、知识点语义关系识别、特征矩阵计算、认知状态评估四大关键技术。之后,文章采用算法模型对比实验、实际应用对比实验,分别对表征认知模型的学科知识图谱和融入学科知识图谱的认知测量模型进行了实验验证,结果表明:ICDA-SKG方法具有有效性和实用性。文章的研究成果可为智能化认知诊断评估提供新思路,并为教育评价实践提供方法指导。

Cognitive diagnostic assessment is an important focus of educational evaluation reform in the new era.However,the existing cognitive diagnostic assessment methods are faced with the problems of low constructional efficiency and poor interpretability in the cognitive model,and at the same time,the cognitive measurement model also has low accuracy due to the lack of semantic cognitive models.Therefore,an intelligent cognitive diagnosis assessment method based on subject knowledge graph,named ICDA-SKG,was proposed in this paper,which included two parts of subject knowledge graph and cognitive measurement model,and whose realization involved four key technologies of subject knowledge point extraction,knowledge point semantic relationship recognition,feature matrix calculation,and cognitive state evaluation.After that,this paper used the algorithm model comparison experiment and the practical application comparison experiment to verify the subject knowledge graph representing cognitive model and the cognitive measurement model integrating subject knowledge graph.The results showed that the ICDA-SKG method was effective and practical.The research of this paper could provide new ideas for intelligent cognitive diagnosis assessment,and offer method guidance for educational evaluation practice.