A Survey on Improvising Frequent Pattern Mining Methods

Abstract

In data mining approaches, association rules mining algorithms are promising for actual applications such as marketing strategies, catalog design. However, the association rules mining is essentially based on a database comprised of Boolean attributes. In order to apply a mining algorithm to further various problems, quantitative attributes should also be appropriately dealt. Fuzzy association rules mining approaches are proposed to overcome such disadvantages based on the fuzzy set concept. The fuzzy association rule mining has a good property in terms of the quantization of numerical attributes of a database compared to generalize Boolean association rule mining. So, the mining results of fuzzy association rules are easy to understand for corresponding human operators.