An Algorithm And Context Analysis Of Mining Frequent Closet Itemsets

Data Mining is defined as a large number of structured and unstructured data to extract useful and meaningful information and knowledge process.With the deepening of the research and development,data mining has been widely applied to a variety of areas.Association rules is an earlier,meaningful research topic in data mining,it is mainly used to found hide,interest and implied links between transactions in large database,generated In the process of generate association rules,mining frequent patterns is the first step,how effective mining frequent pattern has been the focus of attention is to study.In large databases,although the use of an effective algorithm can effectively improve the efficiency of frequent pattern mining,but with the lower of the support,the number of frequent patterns will gradually increase,some of which may be frequent pattern is no useful for user,can not provide effective information for the next step.In this paper,the main reasearch is as follows:1.In-depth study two different strategies of frequent pattern mining methods.The breadth of the search strategy and the depth-first strategy.Reference Frequent Closed Itemsets Mining Algorithm Closet,Closet+ and FP-Close,the paper design a ming algorithm of FP-tree-based frequent closed itemsets FP-FCI.The algorithm is a depth-first algorithm.Analysis and experimental comparison shows that FP-FCI algorithm consumes less time and space in different degrees for data-intensive.2.Analysis the contents of frequent closet items based on the mining frequent closet items result.Design an algorithm HMFP for context analysis based on frequent closet items.In order to better explanation the content of frequent closet items designed a methods for extracting the representative of transactions..