Research And Application On Algorithm Of Association Rules In Commercial Bank Customer Management System

With the mature of database technology and the popularity of data applications, largeamount of data’s produce and collection lead to the explosive growth trend of theinformation resources. Traditional analysis is unable to meet the needs of the managementon data processing. Problems currently facing is how to effectively use these largeamounts of data. We urgently need to find the technology to analyze large amount of data.Facing the challenge, data mining technology is one of the reasonable and useful tools. Inthe area of data mining, the research of association rule receives great attention. Thisthesis focuses on the research of mining algorithm about the association rules. Miningassociation rule can discover the intrinsic link between the individual properties fromanalyzing large and complex data, getting knowledge and rules which have potential valueon decision-making.The drawback of Apriori algorithm is the number of scanning database which mayproduce redundant itemsets. This thesis presents the improved algorithm that databaseonly scan once, saving the time and improving the efficiency of the algorithm executionby learning from affairs compression of Apriori Tids and reducing database scanning trips.The improved algorithm is applied on the management data of the bank customer, such asdata conversion, data integration and data cleaning of32,138pieces of real data of6citiesin one province. This thesis draws some useful rules about applying algorithm to data bypre-processing the data.