基于自适应遗传算法的自动组卷策略研究  

Research on Test Paper Auto-generation Strategy Based on Adaptive Genetic Algorithm

作　　者：余颖[1,2] 李晓昀[1,2] 左贵启[2] 陈波[1] 罗昭[2] 

机构地区：[1]南华大学计算机科学与技术学院,湖南衡阳421001 [2]南华大学船山学院,湖南衡阳421001

出　　处：《南华大学学报（自然科学版）》2016年第3期61-65,共5页Journal of University of South China：Science and Technology

基　　金：南华大学教改基金资助项目(2014XJ-YB13);南华大学船山学院教改基金资助项目(2012CY001)

摘　　要：随着计算机辅助教学和网络的迅速发展,在线考试将逐步替代传统的考试方式而被广泛应用.利用计算机完成自动组卷也将替代原有的人工出卷方式,成为在线考试系统的核心功能之一.本文在量化组卷需求的基础上,提出了一种基于遗传算法的自动组卷策略,将试卷抽象为染色体,每道题抽象为染色体中的一个基因,不断抽取出相应的试题组成多份试卷,最终根据用户设置的组卷参数,选择与用户期望值最接近的一份试卷,本文选取试题号作为单个基因编号对染色体进行编码,避免了解码和多余空间搜索的时间消耗,实验表明基于遗传算法的自动组卷结果较符合出卷要求.With the development of computer-assisted instruction and networking, online examination will gradually replace the traditional exam methods and be used widely. Test paper auto-generating by computer will replace manual way and become one of the core functions of the online examination system.This paper proposed an automatic strategy of test paper generation based on genetic algorithm, taking the quantification of test paper＇ s requirements as precondition.We saw a paper as a chromosome,and each problem as a gene of the chromosome, constantly extracted out corresponding questions to composite several copies of papers, eventually selected the most corresponding copy with the parameters set by user.We chose the number of question as genetic code to encode the chromosome, avoi- ding the time consumption caused by decoding and search of extra space.The result of ex- periment shows that the composition of the test paper is more corresponding to the requirements of user.

关 键 词：遗传算法 自动组卷 在线考试系统 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]