Maintenance of multiple-level association rules for record modification

In the past, researchers usually assumed databases were static and items lay on the same level to simplify the mining problem. Modification of records with item taxonomy is, however, commonly seen in real-world applications. In this paper, we thus attempt to extend Han and Fu´s approach and our previous concept of pre-large itemsets to maintain discovered multiple-level association rules for record modification. The concept of pre-large itemsets is used to reduce the need for rescanning original databases and to save maintenance costs. A pre-large itemset is not truly large, but promises to be large in the future. An algorithm is proposed based on the concept to achieve this purpose. The proposed algorithm doesn´t need to rescan the original database until a number of records have been modified.