Teaching Appraisal' System Based On Data Mining

This article is focused on choosing the target of college teaching appraisal, and the foundation of model on decision-making tree based on teaching appraisal, and the solution of the model. Finally, the author discusses the application of the conclusions concretely.Teaching quality assessment helps to improve the teaching quality and adjust the teaching behaviors. The author creates the data mining system on teaching appraisal, hoping to get helpful information from massive data about teaching and apply them practically.The target system is the foundation of teaching appraisal. So, the author analyzes the model of teaching appraisal system by AHP, and makes clear the assessing targets scientifically. Then, the author analyzes the good and bad points of commonly sued algorithms on data mining, which include Decision-tree algorithm, Cluster algorithm, Apriori algorithm, Genetid algorithm and so on. In view of the massive data about teaching, the author decides to take the decision-making tree algorithm to solve the model. So, the author uses C4.5 algorithm to build the data mining model on teaching appraisal and create a system on it. In data preprocessing, the author defines the threshold values according the system achievements to separate the continuous attributes. First, a decision-tree is built. And then the tree is pruned. Finally, we get the rules by traversing the decision-making tree. The author creates the system by VB development language. There are five modules in this system. By using the system, the author gets 16 rules about teaching evaluation through the data analysis. In all, the system's correct rate arrives 93.33% according the valid test, which illustrates that the function of the system is good.Finally, the author analyzes the functions of the excavation conclusions to the employments of personnel managements, and the possible methods to high teaching quality.