amirhossain shahsavari - department of informatiion technology and computer engineering, azarbaijan shahid madani university tabriz,iran
mohsen mafakheri - department of informatiion technology and computer engineering, azarbaijan shahid madani university tabriz,iran
shahram hosseinzadeh - department of informatiion technology and computer engineering, azarbaijan shahid madani university tabriz,iran
nacer farajzadeh - department of informatiion technology and computer engineering, azarbaijan shahid madani university tabriz,iran

date mining as a tool analyzing date from huge datbaes becomes an important research area. Among date minig tools, association rule ming enables us to find out correlations between date items and also demonstrates which of these correlations are repeated enough, so we can call them strong rules. in this paper we propose a new clustering, based algorithm for mining association rules which removes most of the entire database, finally, we compare our proposed algorithm with the apriori and CBAR algorithms which are two of the most popular algorithms in association rule mining.

Data mining, association rule mining, frequnt iramset mining, CBAR,apriori-like mathods