A Novel Progressive Sampling Based Approach for Effective Mining of Association Rules
Mining Association Rules from huge databases is one of the important issues that need to be addressed. This paper presents a new sampling based association rule mining algorithm that uses a progressive sampling approach based on negative border and Frequent Pattern Growth (FP Growth) algorithm for finding the candidate item sets which ultimately shortens the execution time in generating the candidate item sets. Experimental results reveal that the proposed approach is significantly more efficient than the a priori based sampling approach. Association Rule Mining finds relations among data items in transactions of a huge database. Association rules are frequently used by retail stores to assist in marketing, advertising, floor placement and inventory control.