Hide/Show Apps

Privacy and accuracy systems on financial databases

Bilgen, Adna
A statistical database is a collection of data which contains the sensitive information of individuals. They are extensively used for many purposes. Since the system contains sensitive data of individuals, it must be secure to protect every individual in the data set against attackers. In this thesis, we especially work on the privacy of databases which include financial data. We use data perturbation techniques to work the accuracy and privacy balance between the original database and the perturbed one. We test the accuracy of our masked data on selected statistics. We measure the reliability of our system against existing attack techniques. We develop user-friendly software to use data sanitization by using Java and R programming languages.