Research And Application Of Algorithm Of Fast Mining Association Rules In DM

Data mining is the process of abstracting unaware, potential and useful information and knowledge from plentiful, incomplete, noisy, fuzzy and stochastic data, which is deemed to one of a foreland of data mining system and a promising cross-subject. Association rule is one of more important part in data mining, and finding frequent itemsets plays an important role in abstracting association rules. Most of existing algorithms cost much time and space because they need plenty of repetition.In this paper, we propose an efficient algorithm, which uses database SQL to store information of FP-tree instead of establishing FP-tree method and extracts CLOSET and rules through method of reducing attributes based on analyzed other algorithms. It is fit for abstracting long patterns and saves time and space. Simultaneously, we use this idea to discover functional and approximate dependencies, which avoid blind combinations of attributes based on Apriori. In the cases whose functional dependencies' left-hand side are long patterns, the algorithm is exceptional. For abstracting short patterns, we bring forward alternating pattern decomposition algorithm, which call different algorithms alternately and decompose patterns based on estimating the number of frequent itemsets and that of infrequent itemsets. It changes the dataset each pass, and the decomposed patterns are used to build the next dataset. When the number of frequent itemsets is less than that of infrequent itemsets, we can adopt pattern decomposition algorithm using frequent itemsets, which is the other innovation.Our performance study shows that they are efficient using the algorithms above in their adaptive fields. In addition, we use the algorithm into data of educational administration, taxations and telecom, and abstract some rules for assassinating each department to make advantaged decision for themselves.