A Study of Differentially Private Frequent Itemset Mining

Frequent sets play an important role in many Data Mining tasks that try to search interesting patterns from databases, such as association rules, sequences, correlations, episodes, classifiers and clusters. FrequentItemsets Mining (FIM) is the most well-known techniques to extract knowledge from dataset. In this paper differential privacy aims to get means to increase the accuracy of queries from statistical databases while minimizing the chances of identifying its records and itemset. We studied algorithm consists of a preprocessing phase as well as a mining phase. We under seek the applicability of FIM techniques on the MapReduce platform, transaction splitting. We analyzed how differentially private frequent itemset mining of existing system as well.