ABSTRACT

Frequent pattern mining in databases plays a vital role in many data mining tasks like classification, sequential patters, clustering, association rules analysis etc. There are numerous mining algorithms for finding association rules. One of the most common algorithms is Apriori. It is used to mine frequent item sets from large database. It uses the support statistical measure for pruning the frequent items. When the higher minimum support is used to reduce the number of frequent items for rules generation, this algorithm misses some of the bigger combination of significant frequent items. This proposed DMSA (Dynamic Minimum Support Apriori) generates significant frequent items in large database for association rules generation.