基于蚁群算法的组卷策略  

Composing Test Paper Based on Ant Colony Optimization

作　　者：周维柏[1] 李蓉[1] 

机构地区：[1]华南师范大学增城学院,广州511363

出　　处：《微计算机应用》2010年第6期14-19,共6页Microcomputer Applications

基　　金：广东省本科高等教育教学改革项目(BKJG200765)

摘　　要：针对目前组卷系统数量多但真正能满足用户需求的却非常少的问题。提出了一种新型的智能组卷策略,以试题反应理论为理论基础,将题库中的题目加上鉴别度、关联度、难易度、猜测度、作答时间等属性。以蚁群算法自题库中挑出对于不同程度的学生最合适的测试卷。实验表明本文算法能快速有效地完成组卷过程,具有较强的实用性。A new intelligent test paper is presented for solving the problems that test paper System can not really satisfies the user needs. We adopt the item response theory to formulate the test - sheet composition. Every test item of item bank affects the degree of discrimination, degree of association, degree of difficuh, degree of guess, length of test time, and so on. Our proposed approach generates a satisfactory test - sheet satisfying all criteria from a certain item bank by Ant Colony Optimization. Experiment results show that the algorithm can solve the test paper composition problem quickly and effectively, and also has more capability and utility.

关 键 词：智能组卷 试题反应理论 蚁群算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]