due to increasing use of huge databases , mining practical information and useful knowledge from transactions is evolving into an important area. Most data mining methods focuses on relationship among transactions. Many algorithms have been proposed to find association rules in databases with either binary or quantitative attributes. One of these approaches is fuzzy association rules mining. Fuzzy Apriori and its different variations are the popular fuzzy association rule mining (ARM) algorithms available today. Like the crisp version of Apriori, fuzzy Apriori is a very slow and inefficient algorithm for very large datasets. Hence ,in this paper, we introduce a novel technique , called FCT, for mining fuzzy association rules . Existing method discovers fuzzy association rules by scaning the database once, and performing three tasks simultaneously .First , compute the fuzzy supports of candidate 1-itemsets and then generate large 1-itemsets.Second , divide database into multiple cluster tables ,such that transaction with length k , fall into cluster table k.
Third ,builds new structure called CDi , for each cluster table i, such that CDi[A,x]=∑µA(x) ,where x is an item and A denotes linguistic term. Then fuzzy large item sets are generated according to the cluster tables, instead of scanning whole the database. In addition , if CDi[A,x]=0 for cluster i and item i, then for computing the fuzzy support of each candidate item set containing A(x) , scanning this cluster can be ignored. Consequently , we reduce incredible amount of scanning data and therefore the running time of mining algorithm is reduced greatly. Experimental results show our algorithm is many times faster than fuzzy Apriori for the very large real life dataset .