摘要
基于聚类的考试分析可以将成绩特征相同的考生聚集为一类,从而更加客观和准确地揭示考生的知识和能力水平。将试卷中的各个试题分别视为一个特征项,利用传递闭包法进行模糊聚类分析。在取得聚类结果的基础上,利用Excel对聚类特征项进行再分析,即可得到聚类依据并验证聚类结果。仿真实验表明,基于模糊聚类的考试分析结果明确、可信、有效,并可得到传统的试卷分析难以发现的现象和规律,对于个性化教学和素质教育有很大指导意义,值得推广应用。
Examination analysis based on cluster can cluster the testees with same feature of score to one class, and bring to light the level of knowledge and capability of testees more objectivity and more correctly. The way is to looked upon each examination question as a feature item, and make use of transitive closure for fuzzy clustering analysis. Based on the reselt of the clustering, use the Excel to analysis the clustering feature item farther, to get the reason of the clustering and to validate the reselt of clustering. The simulation show that reselt of examination analysis based on clustering is clear, believable and effective, and can get the hard-won phenomena and rule what by traditional examination analysis. It has important meaning for individuation teaching and education for all-around development, and it is extending and applying worthwhile.
出处
《电脑知识与技术》
2009年第11X期9579-9580,9590,共3页
Computer Knowledge and Technology
基金
河南省教育厅自然科学研究指导计划资助项目(2008B520003)
河南大学教改工程项目(第5批)
关键词
模糊聚类
传递闭包
特征分析
考试分析
fuzzy clustering
transitive closure
feature analysis
examination analysis
