A Study On Association Rules Mining Algorithm And Its Application On Web Mining

Data mining has been becoming more and more popular in the past few years due to the growing demands of database application and the advances in computer technology. Data Mining merges many important research fields including machine learning, artificial intelligent, statistics, knowledge-base systems and data visualization, etc. However, current algorithms proposed for date mining of association rules require several passes over the analyzed database. The I/O overhead in scanning the large database can be extremely high.An efficient algorithm QAIS is proposed that uses the efficient method to reduce database access activity, and present a novel algorithm AIU based on this algorithm, it is fit for mining association rules and incremental updating. It is especially effective in VLDB, mining long patterns, and high support. The Perfermance of QAIS/AIU is verified on the basis of synthetic data, experiments show that the proposed algorithm can mine association rules more efficiently by not generating candidate itemsets and reducing the redundancy of frequent itemsets while generating association rules.And then, a extent of association rules- Time Series Pattern is discussed, and an efficient improved algorithm of web mining is presented. The Performance of algorithms is verified on the basis of synthetic data too.At last, the problems faced in the mining association rules field are discussed in the paper.