Education Information, Association Rule Mining

A large number data have been created in the course of the education informationality. There is a lot of important information about education in this data. We can't find the relation and rules existed in the data because lacked the tools of knowledge discovered. We are unable to forecast future according to the existing data. It caused "the data explosion but knowledge deficient". The data mining means the process that discovered the useful information and knowledge from the data that is massive, not complete, with noise, fuzzy and stochastic. The association rule mining is an important field of data mining. It provides the degree of relevance of the things. There are many association rules hi education data. Data mining can discover it.The association rules express the relationality between the attributes of databases. The rule X=Y, support=s%, confidence=c% means that obtain the Y form X have support=s% and confidence=c%. In this paper we mine the association rules from education information. The rules such as physics=mathematics, support=25% and confidenee=77% can be obtained. Educational activity will be organized and designed scientifically by long time plan according to the rules that express the educational rules, the development of students, and the relevant between the courses.The essential method of association rules mining is based on the large itemsets Apriori algorithm. First, find the frequent itemsets according to smallest support assigned. Second, find the strong association rules according to the smallest confidence. We prefer the MS SQL Server as mining tool among several data mining software. Some modification has been made in Apriori algorithm in the light of particularity of education information. The data have been clustered before to be mined used Apriori algorithm.The main application fields of Education Information Data Mining (EIDM) model are education system. It has characteristic of interactionally, selectivity of source data, integrationally and operationally. The transformation and quantificationmethod manifest the function of processing education data. It is easy to expand function of the EIDM by increasing mining pattern, transformation method andquantification method.In the system designed, the association rules of several application level have been obtain with some different data mining method and parameter. The results can be divided into three kinds according to which they are accepted. The first kind: the rules descript the facts that consist with the understanding about education theory. This association rules obtained by data mining have improved the education theory with data. The second kind: the rule is the fact that has not been realized or paid attention before, but it is recognized the new rules of education and new knowledge now. The third kind: the rule is not discovered before and is not realized by specialist and educationalist now. But it reflected a sort of phenomenon that it has been and it is representative in the certain condition. The reason of this land of rule should be study further.