Incremental update for fuzzy association rules

In practice, many databases are not static but changeable. When some records are added or deleted in a database, it could potentially invalidate some existing rules or introduce new rules. Thus, maintenance of the association rules is an important problem. An incremental update algorithm for the fuzzy association rules is presented; this algorithm uses the last mining results to maintain the collection of frequent fuzzy attribute sets and its negative border along with their support count incrementally. The experimental results show that the incremental update algorithm can effectively save the time due to avoiding the repeated cost introduced by doing mining task directly in the whole database.