Research And Realization Of An Intelligent System Of Test Paper Composition

The test paper generating is an optimized problem to Muti-objective parameter with certain restriction. The optimization is implemented very difficulty by traditional method. The quality and efficiency of auto-generating test paper is all determined by the test-database designs and get problems-terms algorithm. The aim of the research is to design an algorithm that can get a group of test problems terms quickly, which fit the restriction of test requirement. After reading and analyzing a great deal of articles from inside and outside, Genetic Algorithm is selected as the way to produce test paper, and a general-purpose system of auto-generating test paper is designed. The system based on General Genetic Algorithm is designed to search the best answer according to such restriction conditions, like test question types, terms scalar, teaching time, difficulty degree, distinguish and answer time and so on. Finally, it can find the best answer and get ideal test paper.In this search, traditional test generate steps is analyzed, and suitable mathematics model about generating test paper is formed, and the construction of test questions database based on education measure theory and education statistics is built. According to the mathematics model, a producing test paper ideal that the questions are grouped based on their types , count and Proportion has been built. The traditional Genetic Algorithm has been improved at some aspects, and made the response program of producing test-paper through VB6.0' ADO techniques. The testing system not only proved that the article's theory is correct, but also many practical problems in generating test paper have been resolved perfectly, making the improved genetic algorithm is enable in some appliances. The main works of the research can be show as follows:1,Establish a model of generating test paper problem and simplify it.2,Appling ADO techniques, get these question terms from mother database, which is fit the requirements of test paper. After filter, these suitable questions will be put into internal memory. Like this, it enhance the reliability of algorithm.3,Improve the traditional genetic algorithm at Chromosome-Code, replacing the Binary-Code with NaturalNum-Code and directly using the questionID as it. So save the time of Coding and Encoding.4,Applying self-adapting theory to improve the Across-probability and Mutation-probability, it making the Genetic Algorithm can finding the optimizing Across-probability and Mutation-probability to suit itself.5,Contrasting the Small-scalar Selecting Algorithm and K-times Optimized Selecting Algorithm through test, we can proved that the former is more advantage than the later by some test-data.