Research On Algorithm Of Web Mining Based On Fuzzy Association Rules

With the rapid development of information industry, when faced with the multitude of data, people hope that through a deeper analysis of the data in order to dig out the important information that hidden behind a large number of data. However, traditional data management approach is not only unable to discover the link between data and rules, but also can not be based on existing data to forecast the future development trend, which directly led to the "data explosion" and "knowledge poverty" in these two kinds of phenomena simultaneously. As a result, people are forced to find a new method or technology to intelligently, automatically and effectively treated as the existing data into useful information and knowledge. Web Data Mining Technology cames about and gradually developed under this background.This paper analyzes Web data mining technology research status of different mining methods and analysis of their advantages and disadvantages come to Web data mining technology development trends, to understand Web data mining and its related basic concepts and basic skills, in order to subject laid the theoretical foundation for the study.By studying the fuzzy set theory and association rule mining algorithm of the basic concepts, definitions and methods of analysis of some of the existing association rule mining algorithm the advantages and disadvantages for a simple algorithm for mining association rules when carrying out excavation prone excellent demarcation of the boundaries issues, through the introduction of fuzzy set theory and fuzzy set theory, the nature of softening the border, and an existing mining weighted frequent item set algorithm (Mining Weighted Frequent Itemsets, MWFI) based on the improvements proposed mining fuzzy weighted frequent Itemset Algorithm (Mining Fuzzy Weighted Frequent Itemsets, MFWFI).Through the design of experiments to verify the feasibility of improvement, performance of the algorithm MFWFI.The results show that the improved algorithm not only than the original algorithm running time is shorter, and along with the continued increase in the number of records this advantage is even more apparent.