The Design And Implementation Of Information Theory Item Bank And Intelligent Test Paper Generation System Based On Improved Random Sampling Algorithm

With the rapid development of information technology and its deep application in the field of education, item bank management and intelligent test paper generationfor teachers by means of computer technology has become a hot topicof researchand application in the area of the computer aided teaching technology. The use of computer item bank and intelligent test paper generationsystemwhich is beneficial toimprove the quality of teaching can not only reduce the teacher’s time and energy in theitem bank management and test papergeneration instead of heavy tasks, but also can ensure the objectivity and scientific of test paper generation to truly reflect the level of studentlearning and teachers’ teaching effect. Intelligent test paper generationcan be regard asthe optimization problem of multitarget under thespecific constraint conditions, and it is the key to select and improvethe test paper generationalgorithm based on a design.Based on the investigation about researchof item bank and intelligent test paper generation system existingat home and abroad,the authorhaslearned and analyzed the general mathematical modelof test paper generation and test paper parameters, and analyzed the basic process and application of three general test paper generation algorithms. According to the design requirements, this paper have selected and improved the random sampling algorithm. In order to test the feasibility and effectiveness of improved algorithm, the author has done a simulation experimentby comparing the improved random samplingalgorithm with the general random sampling algorithm, which has verified that the improved random algorithm can achieve more knowledge pointcoverage and less item average exposure. According to the full design needs analysis and research of Information Theory, the author has designed and implemented a system of item bank and intelligent test paper generation based on the improved random sampling algorithm, using VB and SQL as the programming language, on the development platform of Microsoft Visual Studio 2008 and SQL Server 2012.The main characteristic of this system is generatinga test paper which has more knowledge pointcoverage and less item average exposure times.More than that,this system can support non-text such as formulas and tables in item, and provides user with some applied function such asusers of the system management, item bank management, optional intelligent or manual wayof test paper generation, item bank statisticsanalysis, TEX automatic layoutand print, and so on. Finally, through the system functional testing, the system can realize the expected functionwell. It has stable operation, friendly interface and good performance of test paper generation, which furtherverifies the feasibility and effectiveness of theimproved random sampling algorithm in this paper.