Design And Realization Of Intelligent Test Paper Auto-generating System Based On Genetic Algorithm

With the development of the computer technology, the way of human’s working ischanging from the heavy manual labor to brainwork. Nowadays, almost all the works inpeople’s life can’t do without the help of computer.At present, online testing system whichis mainly made up of the intelligent grouping test paper has been widely used in all kindsof large-scale level examination of education, in specialized technical personnel andoccupation technical level test. The so-called intelligent grouping test paper is to use thedatabase which has been built to realize computer intelligent selection and test paper. Sothe intelligent test paper is an important measure to realize scientific and normativeexamination, also to realize the separation of the teaching and examination. Thesignificance of research and development of intelligent test paper generation system is thatit not only can save a lot of time but also can eliminate the effect of people’s subjectivewill to make the examination more standardized and objective.By consulting a lot of materials in domestic and overseas, aimed at the problem ofintelligent test paper, this essay analyzed several important indicators and theirrelationship about the examination paper. Based on this, a proper mathematic model wasestablished for test paper system. This system can search quickly according to suchrestriction conditions as test question types, quantities, chapter, answer time, score,difficulty degree and exposition, and then generate a qualified test paper from item bankthrough the selection, crossover and mutation operation, etc. In terms of the groupalgorithm, this paper decided to selecte the genetic algorithm which has the characteristicsof global optimization, convergence speed. By comparing the traditional genetic algorithm,the author designed chromosome coding of genetic algorithm and the choice of the initialpopulation in this system, projected group volume algorithm selection operator, crossoveroperator and mutation operator, and then determined genetic termination conditions of thealgorithm. The experimental results show that the group strategy based on geneticalgorithm can meet the demand of actual test paper generation in efficiency and quality.The system uses SQL Server2005as the background database, C#as a developmentplatform, and adopts ADO technology to connect the database. The system intends to meetteaching and studying demands, above all, to improve working efficiency.Through the application result in a small range, this system operates stably, and it hasgreat applicative value.