Abstract
This paper presents a new hashing algorithm in discovering association rules among large data itemsets. Our approach scans the database once utilizing an enhanced version of priori algorithm, Direct Hashing and Pruning algorithm (DHP). The algorithm computes the frequency of each k itemsets and discovers set of rules from frequent k itemsets. Once the expert in the application domain provides the minimum support, the pruning phase is utilized to minimize the number of k itemsets generated after completing the scanning of specific size database. The required data structure is built to implement the hash table. The analysis shows that the new algorithm does not suffer from the collisions, which lead to high accuracy.
Keywords: Association Rule Mining, Direct Hashing, Basket Market Analysis.