The research and application of general item bank automatic test paper generation based on improved genetic algorithms

Automatic test paper generation is a core stage of computer aided test (CA), is one of the indispensable important technology of computer network test system. This article aim to the goal request of test paper generation, adopt different weight coefficient to different target according to its important, define the fitness function of automatic test paper generation, and use the improved genetic algorithm. The experimental results show that the algorithm can solve the problem of automatic test paper generation compared to other algorithm in item bank very good, and has good performance, high efficiency and quality. At the same time, provide a new method and thought for solving similar multi-objective constraint problems and not neighboring combination problems.