Research On Association Rule Data Mining Algorithm

Data Mining (DM) is a technique that aims to analyze and understand large source data and reveal knowledge hidden in the data. It has been viewed as an important evolution in information processing. During the past decade or over, the concepts and techniques on data mining have been presented, and especially in the latest few years, some of them have been discussed in higher levels.Data Mining uses the classification, association, sequence analysis, clustering analysis, machine self-study and other statistic approaches to find potential, unaware and useful information and knowledge from large database. It is a new subject that involves a lot of subjects and develops with these subjects. Data Mining system can find lots of patterns, in which association rules that describe the interesting relations among the items in given data sets are the important area. This thesis is focus on the correlative study of association rule data mining.Algorithm is key part in DM. On the one hand, Data Mining faces large database, so the efficiency of algorithm is the most important; On the other hand, the computer in use does not meet the demand of the processing of large database. Consequently, we should research and improve present algorithms in order to make them be applied effectively and widely. Based on above, this thesis mainly studies the algorithm of Data Mining.Firstly, this thesis generally discusses the Data Mining, including the concepts and the pattern of the Data Mining, main mining problems, systemic classification, the application and development trend of the Data Mining.Secondly, this thesis deeply researches the Association Rule Algorithm, which is important in the Data Mining. It analyses Apriori algorithm, which is classic one in the Association Rules Algorithms, and the improved algorithms of Apriori, and summarizes problems existing in these algorithms.Thirdly, the thesis introduces in detail DHP (Direct Hashing and Pruning) algorithm and FARM (Fast Association Rule Mining) which is an algorithm improved on the basis of Apriori and DHP.Finally, the author proposes an FARM2 algorithm by analyzing the characteristic and performance of the FARM algorithm, and concludes that the performance of the FARM2 is more effective than those of Apriori, DHP and FARM by comparing the improved FARM with Apriori, DHP and FARM.