Vertical mining for high utility itemsets
Wei Guo Song, Yu Liu, Jinhong Li · 2012
Recently, high utility itemsets mining becomes one of the most important research issues in data mining due to its ability to consider different profit values for every item. In the past studies, most algorithms generate high utility itemsets from a set of transactions in horizontal data format. Inspired by the problem of frequent itemset mining, vertical mining may be a promising approach superior to horizontal mining. In this paper, a high utility itemsets mining algorithm based on vertical database layout is proposed. Candidate high utility itemsets are discovered by intersection of covers at first. Then, high utility itemsets are checked within candidates by scanning database once. Thus, the advantages of vertical database layout, such as low storage, and high efficiency, are utilized. Experimental results show that the proposed algorithm is both efficient and scalable.