It Courses Algorithmic Research And Realization

Computer examination mode applies computer examination system to implementing automatic examination. Computer examination system complete examination work such as generating test paper, examination and test paper scoring. Computer examination mode decreases examination cost and increases examination efficiency. The generating test paper algorithm is one of the key techniques of examination system. The quality of test paper that examination system generates decides whether the construction of examination system is successful or not.This paper researches the application of genetic algorithm in generating test paper. The problem is how to generate the test paper that complies with the test paper requirements. It is in reality a multi-objective optimization problem with multiple constraints. How to ensure that the generated test paper maximum satisfies examination need and has randomness, scientificity and rationality is one of the implementary difficulties. Genetic algorithm is a process searching multi-objective optimization algorithm that emulates the evolutionary process of the nature. Genetic algorithm is executed iteratively to produce evolvement from multiple beginning until the optimal results is generated. According to the characteristics of the test paper problem, this paper solves the test paper problem through improving the standard genetic algorithm. Based on analyzing the basic principle and key technology of the standard genetic algorithm, this paper designs and improves the standard genetic algorithm encoding method, fitness function, selection operator, crossover operator, mutation operator and test paper strategy in terms of the application of the test paper problem. The test paper parameters for genetic algorithm have larger effect on the genetic algorithm performance in relation to the speed and quality of generating test paper. Based on numerous experiments of generating test paper, this paper analyses the experimental results and selects the test paper parameters for this system. The experiments show that the selected parameters has improved the effect of genetic algorithm. This paper compares roulette operator with tournament operator through generating test paper experiments. The experiments show tournament operator has better generating test paper performance than roulette operator.This paper also designs and implements a test question-bank management system for IT courses. The system implements course management, test questions management, automatically generating test paper, test paper management, manually editing test paper and other functions. Considering various constraints such as question type score distribution, knowledge points score distribution, difficulty score distribution, cognition score distribution, automatically generating test paper generates higher quality paper quickly by selecting appropriate parameters based on intelligent technology of generating test paper.