A Frequent Itemsets Data Mining Algorithm Based on Differential Privacy

Differential privacy is a new privacy protection technology, which defines a strict and strong privacy protection model, by adding noise data distortion to achieve the purpose of privacy protection. Frequent pattern mining is an important field in data mining, and its purpose is to find frequent patterns in data set, but the content of the model itself, rules, and counting information is likely to lead to leaking sensitive information. This paper presents a frequent item sets mining method based on differential privacy, named DPFM, which adopts the mining strategy combined with Laplace system and index system, realizing the difference privacy under the premise of guaranteeing performance calculation of privacy protection. Experiments demonstrate that the proposed algorithm, DPFM has an advantage in decreasing error rate, and the convergence rate under two indexes is better than TF method.