Bayesian Networks In Remote Examination System

On the basis of reviewing the development of the Computer Aided Test(CAT), this thesis devotes to the application of Bayesian Networks to the modeling of students in the Remote Computer Aided Test System(RCATS). The aim of the research on the CAT is to simulate the course test for students in the real world. To realize the adaptive personal test,the researchers introduced the Al into this field. The typical intelligent CAT consists of student model,selecting strategy,item database,user interface. The student model is the key part of the CAT. In this paper, we apply the bayes networks on the student modeling. Using the advantages of the bayes networks that can coding the casual relationships between the things, we research in modeling the relationships between the knowledge points of some course. The probabilistic information contained in the bayes networks facilitate the combination of feedback data. On the above basis, we give out the realizing approach of a practical RCATS.