Research And Implementation Of Intelligent Paper Organization Examination System Based On Genetic Algorithm

Under the background of the computer popularization and the rapid computer network development, the paperless examination is inevitable trend development of information technology. Smart group volume and automatic grading not only make the examination more objective and fair, but also reduce teachers' workload in the test process, which save a lot of manpower and material resources. Online examination system makes full use of Web database storage technology, compared to the traditional way that paper-based test examination, it shows its prominent advantages at any time and in unlimited cyberspace. Smart group volume makes the knowledge put forward by expert be together, and extracts the test automatically from the question bank by artificial intelligent way, so form a complete test paper. The intelligent realization of online exam system is the key in the design process. Due to the common volume has slow test reaction, and time-consuming long shortcomings, which reduce the group efficiency. In order to improve the efficiency of the group and quickly generate the group meet the customers' requirements, an optimization group volume scheme based on genetic algorithm has be proposed, which is applied to the actual online examination system.In this paper, theoretical research and development work are as follows:Firstly, analysis of online examination technology about smart group, discusses several typical group algorithm included the working principle and the existing problems in actual application. What's more, analyzes the basic framework of the genetic algorithm, detailed work process and its advantage.Secondly, on the intelligent group system, according to the test item difficulty, knowledge covers and the specific requirements, setting reasonable attribute including the topic and content distribution, and average score, which ensure that can automatically extract a satisfactory paper from the question bank.Thirdly, in order to prevent premature convergence, traditional genetic algorithm is made further improvement work: Use the grouping real number coding scheme in coding genes, and randomly generate initial population by topic for the unit; And will roll into objective function for the optimal solution of mathematical problems; In the operation of choice, to slightly optimization of roulette way, and join the elite reservation methods, which reduce the fine individual discard rate; Crossover operation is also the way of using a separate section to operate,individual topic on variation in different type of segment random operation; In length coding,population size, crossover probability, mutation probability and termination of the offspring and the operation parameters on the set of reference first generally experience a given reference range to determine the initial value, and then continue testing method of the effect of the contrastgroup eventually determine the optimal parameter combination, maximize prevent trapped in local optimal solution.Fourthly, according to the basic design rules of online examination system, the user's specific function requirement, system network architecture, and then the basic framework of the system database is designed. The E-R diagram of the test system is given at the same time, and the user information table and test question type table are set up, database table is optimized.With reference to the common principle of system development, online examination system,which include implemented users login, curriculum management, question bank management,online test, and other functions has be designed.