Research And Application Of Association Rules Mining Algorithm

The emergence of The Data Mining is attributed to the necessary consequence ofthe conflicting movement between the rapid-increasing data and the poor informationday by day. During the past decade or over, produced many new concepts and methodsof Data Mining, and still went further development, Association rule is an importantpattern in Data Mining．Apriori algorithm is the most classic one, but it has somedeficiencies, including scanning databases many times, generating a large number ofcandidate anthology.This dissertation deeply, roundly studies and analyses the Data Mining technique,especially the one for association rules. Based on the basic concepts of the associationrules, analyses the Apriori algorithm of association rules in detail. Systematicallysummarize and analyse the improvement which promote the algorithms efficiency.Finally, we raise the new association rules algorithms to solute some problems. First, itimproves the association rules measures to the shortcomings and the insufficiency ofApriori. And raise new Interesting function, which has high correlations and reduce thegeneration of redundant rules. Secondly, on the basis of the research of Apriorialgorithm, the Apriori algorithm of matrix and two dimensions Hash algorithm isproposed. This algorithm uses the thoughts of hash and matrix just needs scan thedatabase only one time to get the frequent itemsets, it increase the efficiency ofalgorithm.Finally will be connected the association rule mining algorithm to utilize in theteaching appraisal，carries on the Data Mining from the teaching appraisal data. Andfind out the effects of different situation of teaching,helps the teaching department doesthe teaching work well，and improves the quality of teaching.