Clustering-based Multidimensional Sequence Data Anonymization.
Morvarid Sehatkar, Stan Matwin · EDBT/ICDT Workshops · 2014
Sequence data mining has many interesting applications in a large number of domains including finance, medicine, and business. However, Sequence data often contains sensitive information about individuals and improper release and usage of this data may lead to privacy violation. In this paper, we study the privacy issues in publishing multidimensional sequence data. We propose an anonymization algorithm, using hierarchical clustering and sequence alignment techniques, which is capable of preventing both identity disclosure and sensitive information inference. The empirical results show that our approach can effectively preserve data utility as much as possible, while preserving privacy.