遗传算法在试题库智能组卷中的应用

On genetic algorithms in intelligent system of test paper autogenerating from item bank

智能组卷是现代智能教学系统中一个非常重要的课题。以考查点、难度系数和题型为主要控制参数建立了组卷问题的数学模型,并给出了用遗传算法解决组卷问题的新方法。重点阐述了组卷问题的染色体编码方法、适应度函数和遗传算子的设计与实现。实验结果表明,所设计的组卷方法性能好、效率高,而且能确保在一份试卷中不出现考查点重复的试题,是一种实用、有效的组卷方法。

Autogenerating test paper is a very important subject in the modern intelligent teaching system. This paper presented a mathematical model with the main controlling parameters of the main points of knowledge, coefficient of difficulty and the structure of test question. Then it devised a new method of autogenerating test paper by using Genetic Algorithms. The paper focused on description of the encoding method of chromsome, and the design and realization of fitness function and genetic operators. The results of the experiments showed that the method is of high quality and high efficiency. It can avoid repetition of items in a test paper. It is an applicable method with satisfactory performance.