Workshop on Data Mining for Counter Terrorism and Security

Data mining is playing an increasingly important role in sifting through large amount of data for homeland defense applications. However, we must pay attention to the privacy issues while mining the data. This has resulted in the development of several privacy-preserving data mining techniques. The random value distortion technique is one among them. It attempts to hide the sensitive data by randomly modifying the values. This paper questions the utility of the random value distortion technique. The paper develops a random matrix-based spectral filtering technique to retrieve original data from the dataset distorted by adding random values. The proposed method works by comparing the spectrum generated from the observed data with that of random matrices. The paper presents the theoretical foundation and extensive experimental results to demonstrate that the random value distortion technique may not preserve any data privacy after all.