Mining Rare Itemsets on Multi-Level Hierarchies Database

Mining multi-level association rule is one of the most important fields in data mining to exploit hidden knowledge in large database which has extraordinary relations as hierarchies. Unfortunately, with enormous mining outcomes from multi-level hierarchies database show unexpected characteristic that many itemsets, especially leaf level itemsets, are neglected. In this paper, I propose SML algorithm to extract the rare itemsets in any levels of concept in relational database. Statistic data has been used to evaluate and the new proposed information, RareSupp, is used for representing the interestingness that applicable in real-world applications.