Based On Association Rule Data Mining Algorithm

With the development of the updated information technology,especially the emerging of the network technology,our abilities to collect,store and transfer data have been improved dramatically.However,the valuable knowledge or information for decision making is scarce.Facing the vast expanse of ocean of information,people are calling for a technique which can discard the dross and select the essential,discard the false and retain the truth,transfer the data into knowledge,data mining is generated in the background.In fact,the nature of data mining is the process of withdrawing useful and unknown information from incomplete,large fuzzy,and noise practical random data.There are many algorithms concerning about data mining,the most important and traditional method among these is association rule mining algorithm.Hence,it is mainly discussed in the article is about the approved methods of association rule mining algorithm. It includes the following sections:1,Introduce the development and research of data mining at home and broad.2,Highlights the mining association rules theory,first introduced the most basic algorithm—Apriori, as in the real world,different items have different importance,this algorithm introduced the conception of weights.With the rapid development of network,it introduced the conception of parallel.Then with the combination of parallel and weights, proposed a new algorithm—Weighted association rules,which created significant improvement based on Apriori.3,Introduced the concept of granular computing,and proposed a new mining algorithm which combined with granular computing,complete graph,and association rules algorithm.This new algorithm was verified by examples.4,Through the combination of data mining,data warehouse,grain computing etc.We analyse the mini supermarket deeply.Using GRC_G algorithm and weighted association rules to analyse the data association in the mini supermarket.Through the comparison with traditional Apriori algorithm,the performance of the new algorithm was verified.