Skip to Main Content
Mining frequent patterns is a fundamental and crucial task in data-mining problems. This paper proposes a novel and simple approach, which does not belong to the candidate generation-and-test approach (for example, the a priori algorithm) and the pattern-growth approach (such as the FP-growth algorithm) two approaches. This approach treats the database as a stream of data and finds the frequent patterns by scanning the database only once. Two versions of the approach (i.e., mapping-table and transformation-function) are provided. Analyses and simulations of the approach are also performed. Analyses show that the transformation-function version is much better than the a priori and FP-growth ones in storage complexity. Simulation results show that the mapping-table version is comparable to the FP-growth algorithm in execution time.