A Weighted Frequent Itemsets Mining Algorithm Based on Perpendicular Data Format

Mining frequent item sets from a large dense type database may generate a large number of frequent item sets, and it may generate redundant information in some cases. To address these problems, a weighted frequent item sets mining algorithm based on perpendicular data format is proposed in this paper. The algorithm uses constrains of support and weight together, and then by uses item sets extension to mine weighted frequent item sets which meet the support and weight constraints at the same time. In order to reduce the number of candidate item sets, the algorithm used two methods, the first is pruned using property of weighted effectively extension, and the second is use hash table to store weighted non frequent binomial set.