Find Association Rules In Data Mining

With the development of database technologies, people can store data easier. Commerce organization and individual prefer to store daily data, which can be used when needed. Much information is hidden in the vast stored data, people cannot understand these data that appear to be no rule, and people want to utilize these data well by analyzing them in higher level.DM come into being under this background and become to be the study focus in computer field.Firstly, we describe some basic concepts and essence of Data Mining. Then we briefly introduce the study background and study contends of Data Mining, the history and future of Data Mining. Meanwhile, the applications of DM are also introduced.Secondly, we introduce recovering the association rules in DM, which includes its basic concepts. We also analyze the theroy of recovering the association rules and its practical applications.The central topic, which involves Apriori algorithms and its improved algorithms of recovering the association rules in DM. Then we discuss the Apriori algorithms. By our experiments, we found that some shortcomings exist in the connection step and the pruning step of the algorithms and it affect the efficiency of the algorithms greatly. According to the problems, an improved method GPA (Group Parallel Algorithms) was proposed in this thesis. By dynamic grouping, GPA improves the loop efficiency. In addition, we propose an improved algorithm GIUA (Group Incremental Updating Algorithms), which could use the existed results most efficiently. The two improved algorithms could be implemented parallel by using dynamic grouping. The results of our algorithms are also presented in the end of this thesis.To give the difference between the algorithms that gived in this thesis and the same kind algorithms, special compare between GUIA algorithms, GPA algorithms and the same kind algorithms at algorithms' s topic and its excution' s efficiency also presented in this thesis.Finally we draw the general conclusion of this thesis. Future work plan on this specific topic is also included.