Privacy protection of enterprise information through inference analysis

Ensuring that disclosure of information to outside entities is in conformance with the enterprise privacy policies is of utmost concern for all enterprises dealing with consumer information. The existing protection measures proposed for meeting this goal are inadequate. In this paper we present an approach in which the privacy label taxonomy is developed to classify information types in an enterprise by their privacy labels. Inference analysis is performed on the information types using a disjunctive logic programming technique to detect violations of privacy labeling semantics in various information types. The analysis also provides the technique to deal with such violations so as to achieve a violation-free privacy labeling scheme.