In　order　to　make　all　the　association　rule　algorithms　be　used　for　bi-clustering,　the　bi-cluster　problem　has　been　translated　into　the　problem　of　finding　frequent　sets　in　the　association　rule.　With　the　advent　of　many　algorithms　for　bi-clustering,　the　author＇s　technique　not　only　can　obtain　the　bi-clusters,　but　also　can　get　extra　information　associated　with　the　bi-cluster.　Experimental　results　on　gene　expression　data　show　the　effectiveness　of　the　strategy.