Privacy Preserving k-Anonymity for Re-publication of Incremental Datasets

Most of the previous works on k-anonymization focused on one-time release of data. However, data is often released continuously to serve various information purposes in reality. The purpose of this study is to develop an effective solution for the re-publication of incremental datasets. First, we analyze several possible generalizations in the anonymization for incremental updates and propose an important monotonic generalization principle that effectively prevents privacy breach in re-publication. Based on the monotonic generalization principle, we then propose a partitioning based algorithm for re-publication, which can securely anonymize a continuously growing dataset in an efficient manner while assuring high data quality. The effectiveness of our approach is confirmed by extensive experiments with real data.