Inference-usability confinement by maintaining inference-proof views of an information system

  • Authors:
  • Joachim Biskup

  • Affiliations:
  • Technische Universität Dortmund, D-44221 Dortmund, Germany

  • Venue:
  • International Journal of Computational Science and Engineering
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

Extending traditional access control and complementing emerging usage control, inference-usability confinement aims at customising sensitive data to be returned to a client in such a way that the manipulated items are still useful for the recipient but do not enable any usage beyond the intended ones. In the context of a logic-oriented information system, a confinement mechanism generates an inference-proof view of the actually stored instance(s) while interacting with a client. We survey our specific approach to policy-driven inference-usability confinement for a server-client architecture, discussing various parameters and the resulting confinement mechanisms. Basically, the confinement is achieved by enforcing an invariant of the following kind: at any point in time, the information content of the data available to a client does not violate any protection requirement expressed by a declarative confidentiality policy. In this context, the information content of data and, accordingly, the inference-proofness of such data crucially depend on the client's a priori knowledge, general reasoning capabilities and awareness of the confinement mechanism.