Privacy Preserving Closed Frequent Pattern Mining

Mining closed frequent item sets is one of the important problems in data mining. There exists a possibility of designing differentially private Frequent Itemset Mining (FIM) algorithm which can achieve high data utility, efficiency and high degree of privacy. Private Frequent Pattern mining algorithms have a preprocessing phase and mining phase. In the preprocessing phase a novel smart splitting algorithm is used for transforming the database. In the mining phase transaction splitting is done. Certain amount of noise is added to the output for enhancing privacy. The amount of noise added is considerably reduced.