Design And Implementation Of Test Paper Generation System Based On Improved Genetic Algorithm

The traditional papers,is not easy to be randomly generated and has high degree on similarity and difficult to grasp the difficulty coefficient,meantime it is with incomplete coverage on knowledge points,also tedious and time-consuming in the production process.With the rapid development of information technology,intelligent test paper system has become more and more popular,test questions can be shared,random generation of papers,the difficulty coefficient and the knowledge point coverage is controllable,and the generation of papers is efficient and practical.The traditional manual test consumes a lot of time,the quality cannot be assured and knowledge is not comprehensive,the coefficient of difficulty cannot be grasped,leading to the examination with an unsatisfactory result.Since 1975,Professor Holland proposed genetic algorithm,the algorithm has been further improved,and transplanted into the intelligent system of test paper successfully.Using the existing technologies on test paper,to design of a complete set of intelligent test system,gather the wisdom of the teachers group and write high-quality questions,decentralize the individual teacher of the pressure and solve the problem of uncontrollable degree of difficulty and low coverage on knowledge point.This study on this subject is of great significance.The core problem of intelligent test paper system is to solve the problem of calculating and analyzing the initial test paper under multiple constraints and finding out the best set of papers.The whole process of calculation is to solve a series of problems,such as the time efficiency of the test paper,the difficulty coefficient of the test paper,the knowledge point covering problem,etc..In order to solve the above problems,this paper proposes an adaptive genetic algorithm which can generate the test papers in line with the needs of users.In the improved genetic algorithm,taking real segmented in the encoding stage,take the mechanism of roulette selection in the selection stage,the crossover operator and mutation operator to improve the crossover probability and mutation probability according to its fitness value adjusted,to avoid excessive prematurity and slow convergence speed of defect.Finally,in the above technical conditions,the analysis meets the requirements and design of general intelligent test paper system.In this system,it consists of four modules:basic data management,test management,test paper generation module and export test module.Through the system design,the effectiveness of the improved algorithm is further verified.Through the improvement of genetic algorithm,the intelligent test paper system can be a very good practicality,and has a positive impact on society and a great significance.