An Integrated MFFP-tree Algorithm for Mining Global Fuzzy Rules from Distributed Databases

In the past, many algorithms have been proposed for mining association rules from binary databases. Transactions with quantitative values are, however, also commonly seen in real-world applications. Each transaction in a quantitative database consists of items with their purchased quantities. The multiple fuzzy frequent pattern tree (MFFP-tree) algorithm was thus designed to handle a quantitative database for efficiently mining complete fuzzy frequent itemsets. It however, only processes a database for mining the desired rules. In this paper, we propose an integrated MFFP (called iMFFP)-tree algorithm for merging several individual MFFP trees into an integrated one. The proposed iMFFP-tree algorithm firstly handles the fuzzy regions for providing linguistic knowledge for human beings. The integration mechanism of the proposed algorithm thus efficiently and completely moves a branch from one sub-tree to the integrated tree. The proposed approach can derive both global and local fuzzy rules from distributed databases, thus allowing managers to make more significant and flexible decisions. Experimental results also showed the performance of the proposed approach.