A Study On Application Of Data Mining Technology For CET-4Score Analysis

At present the higher education is transformed from elite education to popular educationin china, the school scale is expanding, the number of students is increasing. The data ofstudents has accumulated a lot by using Comprehensive Educational InformationManagement System in universities and colleges. In this paper, we are applying the datamining technology to the performance analysis of the national college English test band4(hereinafter referred to as the CET-4), digging out useful information hidden in theperformance data and providing the basis for improving teaching design and teachingmanagement.This paper mainly includes:1. Mining analysis on college students’ achievements of the CET-4by using the decisiontree algorithm. Establish a decision tree to predict whether students pass CET-4with C4.5algorithm. Cut the branches of the decision tree with post-pruning methods and then extract ofclassification rules. Establish CET-4performance prediction model of students according tothe classification rules and work out the influence degree of various factors on the CET-4grades, providing the scientific basis for improving English instructional design and teachingmanagement and students’ English proficiency and CET-4exam pass rate.2. We use variants of k-means algorithm k-modes for clustering analysis, respectively ofthe CET-4qualified and not qualified, which gives a way to determine the k value for it isdifficult to determine k value. The method chooses the larger k value to cluster, and calculatethe distance between each cluster after clustering, if the distance is below a certain threshold,we think the two clustering is too similar and shuld be combined, then according to the newcombined k value to continue their clustering. We analyze the characteristics of each cluster,and verify the classification rules of decision tree model. So we can divide classes forteaching so as to adopt the method of differential treatment in education, according to theiraptitude, improving the students’ learning efficiency, and pass rate of CET-4.3. The college students’ grade management analysis system was implemented. And thesystem managed entry grades of students, grades of various subjects during the period ofschool, CET-4scores and other academic achievements. The system can also analyze theinformation and results of students and analysis various influence factors in the classificationrules.This paper collected relevant data of students from independent college of localcomprehensive university. Futher analysis the CET-4grades according to the above method.The experimental results verified that the research on the CET-4has played a guiding role.