Khai thác luật phân lớp kết hợp trên cơ sở dữ liệu phân tán

Abstract. This paper proposes a method for mining class association rules in distributed datasets

by using peer-to-peer network. This method utilizes the computing performance of PCs in the network to process the local information at each site, and then only transfers the information of itemsets

whose degree supports satisfy the minimum support threshold to the mining site. Therefore, the proposed method can reduce the memory usage more considerably than that of transferring all dataset’s

information to the mining site.