基于层次分析方法与人工鱼群算法的智能组卷*

针对已有智能组卷方法难以满足考核要求，设计了一种基于层次分析方法与人工鱼群算法的新方法。首先根据试卷总分、考试时间、章节分数、知识点分数、题型分数、能力层次分数、试卷难度与试卷区分度八个考核要求建立多目标、多约束组卷模型；然后利用层次分析方法确定各组卷目标的权重，进而通过线性加权求和将多目标规划模型转换为单目标规划模型；最后利用改进的人工鱼群算法求解模型得到最优组卷方案。该方法适用于单套组卷与多套组卷。同时，讨论了该方法的优缺点以及发展趋势。

Intelligent composing test papers based on analytic hierarchy process and artificial fish swarm algorithm

In view of the shortcomings that satisfy the examination requirements of the existing methods in the intelligent composing test papers, designed a new method based on the analytic hierarchy process and the artificial fish swarm algorithm. Firstly, according to the demands for the test such as the total score, the examination time, the scores of the chapters, the scores of the knowledge points, the scores of the title types, the scores of the ability-levels, the difficult degree and the distinguish degree, constructed a multi-constrained multi-objective programming model of the intelligent composing test papers. Secondly, put the analytic hierarchy process to use in order to work out the weights of all the test objectives. Thirdly, by the linear weighted sum, turned the multi-objective programming model into the single-objective programming model. Finally, with the help of the improved artificial fish swarm algorithm, formed the optimal test paper. The method applied to the composition of a single set of test paper and multiple sets of ones. At the same time, discussed the advantages, shortcomings and development trends of the method.