CAJ | 학술논문

在智能试题库管理系统中，自动生成试卷的科学性和合理性是评价试题库管理系统优劣的关键。自动生成的试卷质量往往取决于试题的属性，如难度、区分度、平均做题时间，等。文章在以往试题库管理系统的基础上，将KNN算法应用到试题库管理系统中，让系统自动确定试题的动态属性。实验结果对比表明，应用KNN算法后，试题的动态属性值更加准确地被预测，从而更好地保证了生成试卷的科学性与合理性，具有重要的应用价值。
재지능시제고관리계통중，자동생성시권적과학성화합이성시평개시제고관리계통우렬적관건。자동생성적시권질량왕왕취결우시제적속성，여난도、구분도、평균주제시간，등。문장재이왕시제고관리계통적기출상，장KNN산법응용도시제고관리계통중，양계통자동학정시제적동태속성。실험결과대비표명，응용KNN산법후，시제적동태속성치경가준학지피예측，종이경호지보증료생성시권적과학성여합이성，구유중요적응용개치。
Scientificity and rationality of automatically-generated test papers are the key to evaluating the quality of the intelligent test library management system . The quality of test papers generated automatically often depends on the properties of test questions , such as difficulty , discrimination and the average question-answering time .In order to determine the dynamic properties of the test questions automatically , this paper , based on the previous test library management system , applies KNN algorithm to the test library management system .Comparison of the test results indicates that the dynamic properties of the questions are more accurately forecast by using KNN algorithm , thus ensuring the scientificity and rationality of generated test papers and having important application value .