Research Of Adaptive Test Paper In The Network Teaching System Based On Genetic Algorithm

The topic is a component of the National Education Scientific Planning Education Examination Scientific Research special topic "intelligent network learning environment construction research of adapting self-examine candidate self-study".With the computer technique is more and more used widely in the field of education and teaching, Computer Adaptive Test have emerged. This method can reflect the level of each examinee more exactly, objectively and reliably, so it has been obtained the recognition of educational circles.Adaptive Test Paper is an important component of Computer Adaptive Test and is a new way of generating test paper. The basic idea is"suit to students in accordance of themselves, examine students in accordance of their aptitude". Adaptive Test Paper refers to combining the technique of AI, the experience of educational expert and the practical level of individual in the process of network teaching, the computer design the content of paper which can adapt with individual special characteristic level. Adaptive Test Paper is a more advanced process of measuring network teaching, it can achieve the purpose of measuring the level of educatees and also is a good way of making educatees to learn and test independently.Based on analyzing of a great deal of domestic and international references, the paper analyzes and evaluates the relation between parameters of the paper and the ability of the examinee, according to the ability of educatee, drew out a copy of test questions which can fit the state of educatee best by making use of the intelligent search and global optimum speciality of Genetic Algorithm.The paper used MATLAB2006, ASP.NET and SQL sever2000 to analyze and design the adaptive test paper. The study will be useful to construct intelligent network learning environment.