A Low-Scan Incremental Association Rule Maintenance Method Based on the Apriori Property

As new transactions update data sources and subsequently the data warehouse, the previously discovered association rules in the old database may no longer be interesting rules in the new database. Furthermore, some new interesting rules may appear in the new database. This paper presents a new algorithm for e ciently maintaining discovered association rules in the updated database, which starts by computing the high n level large itemsets in the new database using the available high n level large itemsets in the old database. Some parts of the n ;1 n ;2 : : : 1 level large itemsets can then be quickly generated by applying the apriori property, thereby avoiding the overhead of calculating many lower level large itemsets that involve huge table scans.

• Type: Conference
• Year: 2001
• Where: AI
• Authors: Zequn Zhou, C. I. Ezeife
• Keywords: AI 2001, Artificial Intelligence, Interesting Rules, Level Large Itemsets, Old Database