An Overview of Traditional Multiplicative Data Perturbation
A Statistical DataBase (SDB) is a database system that allows its users to retrieve aggregate statistics (e.g., sample mean and variance) for a subset of the entities represented in the database and prevents the collection of information on specific individuals. In the statistics community, there has been extensive research on the problem of securing SDBs against disclosure of confidential information. This is generally referred to as statistical disclosure control. Statistical disclosure control approaches suggested in the literature are classified into four general groups: conceptual, query restriction, output perturbation and data perturbation. Conceptual approach provides a framework for better understanding and investigating the security problem of statistical database at the conceptual data model level.