Abstract:

The issue of data mining for finding frequent itemset in a dataset is present in the literature for over two decades. During this time many algorithms for frequent itemset mining were introduced, the most popular of which is the Apriori algorithm. In this article an improvement for Apriori algorithm is proposed, that is based on items grouping. The proposed Heuristic Grouping Algorithm (HGA), depending on the selected parameters, can be characterized by a significant reduction in computational complexity compared to the original Apriori algorithm.