An Application Research Of Data Ming Technology In Higher Educational Management System

With the rapid development of information technology and continuous growth of higher educational scale,information technology has become an inevitable choice for the modernization of higher education.Educational Management Information System which is based on the network has been widely used and promoted in all colleges and universities,it helps promoting the innovation of higher education management system and improving the management level of teaching management department.Educational Management System has achieved managing the daily teaching and accumulated a large amount of useful data resources.How to use modern technology to process these huge useful data sources and accurately discover the hidden knowledge is not an easy task.The data mining techniques which is recently emerged has provided a strong technical support for solving these problems.Using data mining techniques to mine useful knowledge or information from these huge and valuable data and then apply these extracted knowledge into teaching management has been significant meaning for managem ent workers to improve the ability of decision-making and management level in educational administration.This article firstly introduced the theory and techniques in data mining,then the process and common methods used in data mining,we particularly fo cus on study of the theory and process of cluster analysis methods,including K-means clustering algorithm(K-Means)and fuzzy C-means clustering algorithm(FCM).Then,one of the most popular recommendation technology,collaborative filtering techniques w hich is applied in our methods is described in detail,including user-based collaborative filtering techniques and Slope One technology.Collaborative filtering technique uses a similar user evaluation of certain information to form the prediction of speci fied user preference information of the system,when it is applied into the student achievement system in Educational Management System,it can perform well for analysis and prediction of the students' achievement.After introduced the clustering technology and collaborative filtering techniques in detail,this article proposes three methods which combined these two technologies to predict the achievement of students,they are prediction method based on K-means clustering and student-based collaborative filtering(KMCF),prediction method based on Fuzzy C-Means and student-based collaborative filtering(FCMCF),prediction methods based on Fuzzy C-Means clustering and Slope One algorithm(FCMSO).Finally,these prediction methods are applied into Educational Management System to analysis and predict the achievement of students,experimental results demonstrates these method can achieve much better results.