Research And Application On Association Rules Mining Method

Association rules as an important branch of data mining research,has been one of the most active research methods.Since the mining methods of association rules can find out the data that can not directly see the relationship from the surface,It has been widely used in retail business,Web mining,text mining and some other fields.With the rapid development of mobile communications and Internet technology,the size of data is also growing which results in that traditional mining association rules can hardly meet the need of large-scale data analysis.So how to improve the association rules to accommodate the needs of a large amount of data has become one of the hottest areas of data mining.In this paper we describe some of the related conceptual and methodological premise of association rules,and elaborate the knowledge and problems of classical Apriori algorithm association rules.The main content of this paper is as follows:First,in order to avoid the problems that traditional association rules Apriori algorithm has to scan the global database many times to generate the frequent item sets.Optimization database encoding rules is proposed in the basis of classical Apriori algorithm.It removes the item which does not meet the minimum support.A judging mark(Judgemark)is added to database to decide whether the transaction database is frequent or not.This mechanism improved the MRS-Apriori algorithm in connecting to the database to scan the database efficiently.Second,on the basis of the encoding rules,traditional Apriori algorithm generate candidate item sets in large number will consume too much time.The algorithm named MRS-Apriori using Hadoop MapReduce programming framework model to achieve parallel processing.It increase the performance of iteration when connecting process efficiency,reduces the time cost of large-scale data operation.The experimental results show that the improved MRS-Apriori algorithm can effectively reduce the computation time,with high accuracy in handling large data sets.Association rules have different problems and challenges in different social development stages.In the future,it will improve the applicability and stability of the algorithm in order to adapt to the future development further.