Incremental Update of Discovered Association Rules
Partitioned derivation and incremental updating.
A fast updating algorithm, FUP (Cheung et al.’96)
- View a database: original DB È incremental db
- A k-itemset (for any k),
* large in DB È db if large in both DB and db
* small in DB È db if small in both DB and db
- For those only large in DB, merge corresp. counts in db
- For those only large in db, search DB to update their itemset counts
Similar methods can be adopted for data removal and update.
Principles are further developed for distributed/parallel mining of association rules.