The Research And The Application To The Exam Bank System Based On The Improved Adaptive Genetic Algorithm

As the rapid development of the information technology, the way of the education has experienced a huge change. And the ways to test are variable. The on-line examination is one of them. This paper is discussed the problem on the test paper intelligent generation system. The efficiency of the system has been improved, by the improved traditional genetic algorithm.The problem of the test paper intelligent generating system is how to deal with the optimization of the multi-objective, under the constraints. The traditional ways, such as random choosing and backtracking, cannot deal with this problem properly. They bring a lot of problems, such as the efficiency of the generating time and the success rate of the paper generating. To deal with these problems, the paper is discussed an improved genetic algorithm, to satisfy the user’s demand.In the improved adaptive genetic algorithm, the method based on the integer segmentations is applied to the coding stage; the competition choosing mechanism is applied to the choosing operation stage. The traditional genetic operator is improved. In the process of the paper generating, the system can change the algorithm adaptively, depending on the operating degree. This way can not only avoid the problems has been brought using tradition algorithm, but also improve the operating efficiency to some degree.On the demand of the test paper intelligent generating in the exam system, we improved the variation of traditional genetic algorithm. During the system running, the chromosome of the operator can be varied adaptively, depending on the weight of the different varied constrains, to improve the guidance of the genetic algorithm.Analyzing the demand of the test system of the basic computer, we designed and achieved the system. The system contains three main function modules, including the test paper managing module, test scoring system and examinee test process managing system. Through the system running, the effect of the improved algorithm has been proved.This algorithm has been applied to the on-line test system. Comparing to the original algorithm, this improvement of the algorithm has been proved to be more effective and better.