Using two threshold for mining erasable itemset on dynamic incremental database

Abstract:

Traditional data mining is often applied in static databases and bath processing. In fact, databases are often changed, bath processing is not suitable because it consumes more time to mine in the whole database. Therefore, mining in dynamic databases attracted many researchers where mining erasable itemsets (EIs) in incremental databases is one of interesting areas. Recent years, there are some publications developed for updated EIs in dynamic databases but they consume more time to rescan the original database. In this paper we propose an algorithm for updating EIs using two minimum thresholds to avoid rescanning databases, we also use new data structure to efficient process incremental databases.