Ensuring authorization privileges for cascading user obligations

  • Authors:
  • Omar Chowdhury;Murillo Pontual;William H. Winsborough;Ting Yu;Keith Irwin;Jianwei Niu

  • Affiliations:
  • The University of Texas at San Antonio, San Antonio, TX, USA;The University of Texas at San Antonio, San Antonio, TX, USA;The University of Texas at San Antonio, San Antonio, TX, USA;North Carolina State University, Raleigh, NC, USA;Winston-Salem State University, Winston-Salem, NC, USA;The University of Texas at San Antonio, San Antonio, TX, USA

  • Venue:
  • Proceedings of the 17th ACM symposium on Access Control Models and Technologies
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

User obligations are actions that the human users are required to perform in some future time. These are common in many practical access control and privacy and can depend on and affect the authorization state. Consequently, a user can incur an obligation that she is not authorized to perform which may hamper the usability of a system. To mitigate this problem, previous work introduced a property of the authorization state, accountability, which requires that all the obligatory actions to be authorized when they are attempted. Although, existing work provides a specific and tractable decision procedure for a variation of the accountability property, it makes a simplified assumption that no cascading obligations may happen, i.e., obligatory actions cannot further incur obligations. This is a strong assumption which reduces the expressive power of past models, and thus cannot support many obligation scenarios in practical security and privacy policies. In this work, we precisely specify the strong accountability property in the presence of cascading obligations and prove that deciding it is NP-hard. We provide for several special yet practical cases of cascading obligations (i.e., repetitive, finite cascading, etc.) a tractable decision procedure for accountability. Our experimental results illustrate that supporting such special cases is feasible in practice.