Analysis And Research Of Student Score Based On Data Warehouse

With the development of information construction of our graduate school, the Graduate Comprehensive Educational System has been put into use. Since only the basic functions of data query and backup have been implemented in the current system and the analysis function is not strong enough, a large amount of student data cannot be utilized. The score of students can be used to measure the quality of students as well as the quality of education. The education quality reflects the competence of a university. Therefore, policymakers have more concern on distribution of students score and factors that contributes to the results in order to improve the quality of education.Therefore, the score of students are analyzed for research. A data warehouse of students score is created. The process contains the establishment of3-layer data models and the cube. Some conclusions are drawn using the pivot table service of EXCEL. For a more in-depth study, the decision tree and association rules analysis method are introduced. After determining the data set, two decision tree models were established based on the algorithm of ID3and C4.5to find the factors of student score, and then verify accuracy, recall, F-value with the test set. Comparing the results of the two algorithms, we find C4.5algorithm has higher accuracy and does the pruning operation during building the tree to make less notes and more compact structure, and it’s rules are more concise and easier to be understood to meet the needs of policymakers. Thus, we draw some conclusions that can be used for student enrollment according to C4.5algorithm.In our training program, the courses are categorized into public courses, basic theory courses and basic professional courses. Among them, public courses are the prerequisites of basic theory courses; basic theory courses are the prerequisites of basic professional courses. The prerequisites courses may have huge influence on the following courses. This thesis implements classical algorithm of association rules analysis-Apriori algorithm to find out the relation between public courses, basic theory courses and basic professional courses. Some predictive conclusions are drawn for enrollment of graduate students.