Frequent closed multi-dimensional multi-level pattern mining

Recently, several efficient frequent closed itemset mining methods have been proposed. Those methods are only able to mine relationship among item information. However, real life colleted data in transaction database usually contain many interesting useful information not only item information but also dimension and hierarchal information. Such information can be used for analysis on knowledge discovery in database system. In this paper, we propose a new method, called frequent closed multi-dimensional multi-level pattern mining, which is suitable for mining frequent patterns in real life information. In additions, we show that 1) our completed frequent closed multidimensional multi-level patterns are smaller than the number of multi-dimensional multi-level frequent patterns and 2) our closed multi-dimensional multi-level patterns represent all patterns of multi-dimensional multi-level in equivalence class with the same support.