Research And Implementation Of Intelligent Test Paper System Based On Improved Genetic Algorithm

This paper makes a deep analysis of the related theory of intelligent test paper generation algorithm which is based on the reference of domestic and foreign literatures.After studying the core issues of intelligent test paper generation,an improved genetic algorithm for intelligent test paper generation is proposed,which is applied and implemented in the test paper generation system.Paper generation itself is a very complex optimization problem for multi-item combination of label constraints.The core of intelligent paper generation is to complete the coding design of the algorithm,which determines the efficiency of paper generation and the quality of the paper.This paper proposes the use of genetic algorithm to realize the intelligent generating test paper system according to the analysis and study of the related problems and the mathematical model of generating test papers.At the same time,this paper is aimed at the problem of premature convergence and low search efficiency of genetic algorithm itself to improve the genetic algorithm which is mainly embodied in the re-optimization of the generating test paper business process.It optimizes and improves proposition test paper generating business process,mutation operator,crossover operator,population size and genetic operator.Finally the intelligent test paper generating system which is based on improved genetic algorithm is realized.Firstly,according to the business requirement and the core difficulty of the system,the paper designs the steps and mathematical model of intelligent test paper generation.Secondly,the test paper generating system based on genetic algorithm is designed,including the improvement of genetic algorithm and the design of test paper generating process.Furthermore,the database of the system and the core business of the system are analyzed and designed.This intelligent test paper system covers four core businesses,including the item bank management,proposition management,test paper management and statistical analysis.It corresponds to the daily workflow of proposition,basic management,automatic test paper generation,test paper management tracking,and feedback of measured data.Finally,the running efficiency and accuracy of the system are tested,and the experimental data are verified.The test results show that the system can automatically generate high-quality test papers that meet the constraints.