The Application Of Decision Tree Algorithm Based On Rough Set In Higher Education Assessment

Higher education assessment is an important part of the management of higher education, and it is an effective way to examine the process of teaching. It not only plays an adjusting, guiding, controlling and promoting role for teaching, but also an important meathod of evaluating the work of teaching.Currently office of academic affairs would feedback the score results or assessment grades to a teacher in the final examination. The method is very effective in education reform and improving the quality of teaching, but the method do not point out what areas need improvement for the teacher. Office of academic affairs can not find hidden rules from the large number of assessment data.Decision tree classification is a regular technology of data mining. It has the features of high-speed and high efficiency. It is not only used to analyze the data, but also used to predict. However the properties of the data set are too many, therefore decision tree are prone to poor decision tree structure. It is difficult to find useful information. Rough set is an effective tool to deal with incomplete information. It is widely used in data mining, data preprocessing and attribute reduction, because it has some advantages in eliminating redundant information and handling large data sets. Because of the classification lack cross validation function the classification are not high accuracy and instability, therefore decision trees and rough sets are highly complementary strengths. In this paper, combination of the rough set theory and the decision tree method will be used in university evaluation of teaching. First, ruducing evaluation indicators by rough sets, then deleting redundant teaching assessment indicators, Secondly, building the decision tree by removing the evaluation of teaching form. The paper presents a decision tree construction algorithm based on attribute importance. The algorithm constructs a decision tree recursively by calculating the importance of properties of the relative division. Thirdly, constructing a decision tree by evaluation of teaching form, and generating decision rules. Finally, explicating the rules which guides the teaching ultimately.