Understanding privacy risk of publishing decision trees
DBSec'10 Proceedings of the 24th annual IFIP WG 11.3 working conference on Data and applications security and privacy
Can the Utility of Anonymized Data be Used for Privacy Breaches?
ACM Transactions on Knowledge Discovery from Data (TKDD)
A heuristic data-sanitization approach based on TF-IDF
IEA/AIE'11 Proceedings of the 24th international conference on Industrial engineering and other applications of applied intelligent systems conference on Modern approaches in applied intelligence - Volume Part I
Using TF-IDF to hide sensitive itemsets
Applied Intelligence
| Hi-index | 0.00 |
Data publishing can provide enormous benefits to the society. However, due to privacy concerns, data cannot be published in their original forms. Two types of data publishing can address the privacy issue: one is to publish the sanitized version of the original data, and the other is to publish the aggregate information from the original data, such as data mining results. There have been extensive studies to understand the privacy consequence in the first approach, but there is not much investigation on the privacy consequence of publishing data mining results, although, it is well believed that publishing data mining results can lead to the disclosure of private information. We propose a systematic method to study the privacy consequence of data mining results. Based on a well-established theory, the principle of maximum entropy, we have developed a method to precisely quantify the privacy risk when data mining results are published. We take the association rule mining as an example in this paper, and demonstrate how we quantify the privacy risk based on the published association rules. We have conducted experiments to evaluate the effectiveness and performance of our method. We have drawn several interesting observations from our experiments.