Research And Implementation Of Knowledge Points Association Analysis And Test Paper Generating Algorithms Of College English Diagnostic Test System

In order to meet the urgent need of the students to improve the scores of CET, we have developed a College English Diagnostic Test System (CEDTS). The test paper generating algorithm of the weakness of knowledge points in the current system doesn’t consider about the associations of the knowledge points, which can’t satisfy the needs of learners well. The research at the early stage had many limitations, the reliability of the association rules generated by it is not high and the corresponding test paper generating algorithm is also relatively rough which cannot be put into the practical application. So this paper researched the knowledge points association analysis of CET by data mining with large amounts of data, put forward the test paper generating algorithm based on knowledge points association rules and implemented to perfect the functions of the system by coding.First, this paper carried on knowledge points association analysis by a large amount of data collected in the practical application, the data mining process included data collection, data preprocessing, the actual mining, results analysis and so on, built the table of the knowledge points association rules. Second, from the existing test paper generating algorithms of the CEDTS, the table of the knowledge points association rules was introduced, the paper improved and implemeted two test paper generating algorithms of the weakness and coverage maximization based on knowledge points association rules. Last, the test paper generating algorithms of question types was researched and discussed, the paper put forward and implemented two new test paper generating modes of according to the weakness of question types and according to the weakness of both question types and knowledge points, we put forward the idea of association analysis of both knowledge points and question types by using the method of knowledge points association analysis, and prototypes of test paper generating algorithms based on association rules of question types and knowledge points. This paper also implemented by coding to perfect the functions of test paper generating in the CEDTS and carried on the related validation and analysis of question recommendation and grade improving.The knowledge points association analysis in this paper is based on large amounts of data collected from practical application, knowledge points association rules table had a large increase in some aspects of the minimum confidence, the minimum support, the lift and maximum number of associated. The two test paper generating algorithms based on knowledge points association rules can both consider about the association between the knowledge points better than the original ones to present the guidance and paper recommendation according to the knowledge points mastering situation of learners practically and effectively. The research and design on aspects of the question types can diagnose the weakness of the question types of the learners, refining the learners’ situation further combining the knowledge point can make the paper recommendation more comprehensive, which has wide application prospect and research value.