A Novel Incremental Updating Algorithm for Maintaining Discovered Negative Association Rules

Recently, mining negative association rules is an important research topic among various data mining problems and has been proved to be useful in real world. The issue of maintaining discovered negative association rules is paid more attention in the same way. Especially, the process of updating frequent negative itemsets is still a complicated issue for dynamic database that involve frequent additions. This paper presents an efficient algorithm INAR for mining negative association rules in incremental updating databases. With a correlation coefficient measure and pruning strategies, the INAR algorithm can find all valid negative association rules quickly and overcome some limitations of the previous mining methods. The experimental results demonstrate its effectiveness and efficiency.