Association Rule Mining Algorithm

With the development of database technology and the diversification of the means for people to obtain data,human beings had a rapid increase in the amount of data,and they were faced with the dilemma that there were rich data but poor knowledge,data mining technology helped the humanity get rid of this impasse.Data mining integrates many disciplines effectively,such as the database,artificial intelligence,mathematical statistics,machine learning and pattern recognition,it can extract information and knowledge that are implicit in the work,not known in advance,understandable and potentially useful for the data owners from a lot of data with noise,and it has a wide range of applications in the retail, telecommunications,financial industry and other fields.Association rule mining is one of the most widely studied topics in the field of data mining. After Agrawal R. and others raised the issue of association rule in 1993,many scholars,have undertaken extensive research on the issue and made a lot of new algorithms.Based on the study and summarization of association rule mining, this paper carried out a detailed description and analysis of classical Apriori algorithm.According to the insufficiencies of Apriori algorithm,it proposed updating problem of association rule,and divided the problem into two categories:Dynamic adjustment of threshold and dynamically updating database,meanwhile,issued updating algorithm for each category, introduced the FUP algorithm and PFUP algorithm in details.Considering the deletion of data and the addition of data in database updating,this paper gave a new updating algorithm,based on the PFUP algorithm this algorithm raised the concept of new frequent itemsets, hence,greatly expanded its application.Since these algorithms are based on Apriori algorithm,while these two shortcomings repeatedly scanning the database and producing a large number of candidates seriously impact the efficiency of Apriori algorithm,based on the ABM and ABBM algorithm this paper proposed algorithm of matrix-based association rule,and set up an intermediary matrix,that provides the basis for resolving updating issue.Finally,concluding the contents of former chapters,it raised a updating algorithm of matrix-based association rule.The algorithm completely put care of the Apriori algorithm models,taking into account both cases of updating issue,without scanning the original database,just scanning updated part of database,realized the updating of association rule efficiently.