基于回归算法的试题质量评价系统研究

A quality evaluation system of examination questions based on the regression algorithm

试卷评价是教育评价的重要部分,现有的试卷评价系统利用统计方法求解出预定指标项的值,并用直方图表示考生总分的分布情况,以此对试卷有一个全面宏观的了解。然而,这些指标缺乏对具体考题进行质量评价,难以作出进一步的分析和评价。文章结合KDD领域中的有关技术,提出一种数据表示形式,在此基础上求解出一系列针对考题及考题集的评价指标,以便更深入地反映试卷的具体情况,这一求解结果也有助于试题库的维护。

The evaluation of test papers is an important part of the evaluation of education. In the current evaluation system of test papers,a series of indexes are analyzed by the method of statistics, and histograms are used to demonstrate the distribution situation of marks,then a macroscopical description of the test papers can be made, but the discussion on the concrete examination questions is lacking. In this paper,a data structure is proposed based on the techniques of knowledge discovery in databases(KDD),and the relationship between the marks and the numbers of students is shown with a two-dimensional table. And a series of evaluation indexes of examination questions are produced to reflect the microcosmic instances of papers,which is helpful to the maintenance of the base of examination questions.