Research On Generating And Analyzing Test Papers In Item Bank System

Computer-Assisted Instruction is one of computer application mode. It is based on the artificial intelligence science, cognitive science, computer science education theory and psychology. It is a new domain of computer application technique, and represents a new education thought and teaching method. Item bank system is one key part of CAI system which can carry through generating test paper and analyzing test paper, and it is an important method to assure test item having better quality, keeping level stabilization and getting better test and diagnoses. According to problems of low intelligence, generating test paper's low efficiency and so on in item bank system, this paper firstly analyses the relation between knowledge point network and test paper based on the research on knowledge point and Relationships between knowledge points. Secondly, this paper analyses and construct multi-constraint problem which should be considered during intelligently generating test paper based on research on manual generating test paper, test theory and knowledge point hiberarchy, builds up math description of single and group test paper, realizes single and group test paper generation using adaptive Genetic algorithm, and it is proved that it has higher efficiency than other algorithms of generating test paper through experiment. Finally, this paper put forwards test paper's analysis in term of knowledge point, which includes building up a model of fuzzy evaluation in term of fuzzy theory and education psychology, which includes objective evaluation on student's test score and teacher's subjective evaluation, so it can evaluate student's test result more reasonably, and builds up vector space model based on student's difficult knowledge points'diagnoses to commend corresponding teaching material to them.