Security of random data perturbation methods
ACM Transactions on Database Systems (TODS)
Privacy-preserving data mining
SIGMOD '00 Proceedings of the 2000 ACM SIGMOD international conference on Management of data
Privacy-preserving k-means clustering over vertically partitioned data
Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining
Privacy-Preserving Distributed Mining of Association Rules on Horizontally Partitioned Data
IEEE Transactions on Knowledge and Data Engineering
IEEE Transactions on Knowledge and Data Engineering
Geometric methods for mining large and possibly private datasets
Geometric methods for mining large and possibly private datasets
An attacker's view of distance preserving maps for privacy preserving data mining
PKDD'06 Proceedings of the 10th European conference on Principle and Practice of Knowledge Discovery in Databases
KES '08 Proceedings of the 12th international conference on Knowledge-Based Intelligent Information and Engineering Systems, Part II
Reconstructing Data Perturbed by Random Projections When the Mixing Matrix Is Known
ECML PKDD '09 Proceedings of the European Conference on Machine Learning and Knowledge Discovery in Databases: Part II
A constraint satisfaction cryptanalysis of bloom filters in private record linkage
PETS'11 Proceedings of the 11th international conference on Privacy enhancing technologies
Non-metric multidimensional scaling for privacy-preserving data clustering
IDEAL'11 Proceedings of the 12th international conference on Intelligent data engineering and automated learning
Breaching Euclidean distance-preserving data perturbation using few known inputs
Data & Knowledge Engineering
Efficient two-party private blocking based on sorted nearest neighborhood clustering
Proceedings of the 22nd ACM international conference on Conference on information & knowledge management
Hi-index | 0.00 |
One of the fundamental challenges that the data mining community faces today is privacy. The question "How are we going to do data mining without violating the privacy of individuals?" is still on the table, and research is being conducted to find efficient methods to do that. Data transformation was previously proposed as one efficient method for privacy preserving data mining when a party needs to out-source the data mining task, or when distributed data mining needs to be performed among multiple parties without each party disclosing its actual data. In this paper we study the safety of distance preserving data transformations proposed for privacy preserving data mining. We show that an adversary can recover the original data values with very high confidence via knowledge of mutual distances between data objects together with the probability distribution from which they are drawn. Experiments conducted on real and synthetic data sets demonstrate the effectiveness of the theoretical results.