基于差分进化算法的英语考试智能组卷

Intelligent Test Paper Composition on English Examination Based on Differential Evolution Algorithm

为保证英语大规模标准化考试质量要求和信息化建设的需要,针对传统组卷算法存在的问题,基于差分进化算法展开研究。首先,在阐明影响试卷质量指标的基础上,构建了组卷算法数学模型;然后,研究差分进化算法模型,主要针对变异、交叉和选择三种操作;最后,以组卷算法和差分进化算法模型为基础,绘制了差分进化算法智能组卷程序流程图,并对主要步骤进行了说明。差分进化算法具有原理简单、受控参数少、鲁棒性强、收敛性和稳定性好等特点,依据文中研究内容可以快速构建适合于大学英语教学和考试的智能组卷系统。

In order to ensure the quality requirements of large- scale standardized tests in English and the needs of information construction,study is carried on based on differential evolution in the light of the problems in traditional test paper algorithm. First,a mathematical model of test paper algorithm was constructed after illustrating indicators effected papers＇ quality. Then,the model of differential evolution was studied,especially for mutation,crossover and choice. Finally,the program flowchart of intelligent test paper composition based on differential evolution was drawn according to test paper algorithm and model of differential evolution algorithm,and the main steps were described. Differential evolution algorithm has many characteristics,such as simple in principle,less controlled parameters,strong robustness,good convergence and stability and so on. Intelligent test paper composition system for college English learning and test can be quickly built based on the content studied.