Privacy Preserving Mining Sequential Pattern based on Random Data Perturbation

Abstract

Privacy preserving data mining is a hot research direction of data mining in the big data environment. If Data mining has been used properly, it will endanger the privacy and the information. Inspired by that, we researched a method via random data perturbation method to keep privacy preserving in sequential patterns mining. Its strategy is as follows: First, noisy events are added into each event sequence of the original database. Then, an algorithm PP-Span (Privacy-Preserving Sequential pattern mining) is proposed to reconstruct the frequent sequences from these noise-added data sequences. Our results showed that it can outperform a lot of traditional methods.