Minimal linear invariants

  • Ming-Yang Kao
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1023)


To protect sensitive information in a cross tabulated table, it is a common practice to suppress some of the cells. A linear combination of the suppressed cells is called a linear invariant if it has a unique feasible value. Because of this uniqueness, the information contained in a linear invariant is not protected. The minimal linear invariants are the most basic units of unprotected information. This paper establishes a fundamental correspondence between minimal linear invariants of a table and minimal edge cuts of a graph constructed from the table. As one of several consequences of this correspondence, a linear-time algorithm is obtained to find a set of minimal linear invariants that completely characterize the linear invariant information contained in individual rows and columns.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    A. V. Aho, J. E. Hopcroft, and J. D. Ullman. The Design and Analysis of Computer Algorithms. Addison-Wesley, Reading, MA, 1974.Google Scholar
  2. 2.
    C. Berge. Graphs. North-Holland, New York, NY, second revised edition, 1985.Google Scholar
  3. 3.
    G. J. Brackstone, L. Chapman, and G. Sande. Protecting the confidentiality of individual statistical records in Canada. In Proceedings of the Conference of the European Statisticians 31st Plenary Session, Geneva, 1983.Google Scholar
  4. 4.
    T. H. Cormen, C. L. Leiserson, and R. L. Rivest. Introduction to Algorithms. MIT Press, Cambridge, MA, 1991.Google Scholar
  5. 5.
    L. H. Cox. Disclosure analysis and cell suppression. In Proceedings of the American Statistical Association, Social Statistics Section, pages 380–382, 1975.Google Scholar
  6. 6.
    L. H. Cox. Suppression methodology in statistics disclosure. In Proceedings of the American Statistical Association, Social Statistics Section, pages 750–755, 1977.Google Scholar
  7. 7.
    L. H. Cox. Automated statistical disclosure control. In Proceedings of the American Statistical Association, Survey Research Method Section, pages 177–182, 1978.Google Scholar
  8. 8.
    L. H. Cox. Suppression methodology and statistical disclosure control. Journal of the American Statistical Association, Theory and Method Section, 75:377–385, 1980.Google Scholar
  9. 9.
    L. H. Cox and G. Sande. Techniques for preserving statistical confidentiality. In Proceedings of the 42nd Session of the International Statistical Institute. the International Association of Survey Statisticians, 1979.Google Scholar
  10. 10.
    D. Denning. Cryptography and Data Security. Addison-Wesley, Reading, MA, 1982.Google Scholar
  11. 11.
    D. Gusfield. A graph theoretic approach to statistical data security. SIAM Journal on Computing, 17:552–571, 1988.Google Scholar
  12. 12.
    M. Y. Kao and D. Gusfield. Efficient detection and protection of information in cross tabulated tables I: Linear invariant test. SIAM Journal on Discrete. Mathematics, 6(3):460–476, 1993.Google Scholar
  13. 13.
    G. Sande. Towards automated disclosure analysis for establishment based statistics. Technical report, Statistics Canada, 1977.Google Scholar
  14. 14.
    G. Sande. A theorem concerning elementary aggregations in simple tables. Technical report, Statistics Canada, 1978.Google Scholar
  15. 15.
    G. Sande. Automated cell suppression to preserve confidentiality of business statistics. Statistical Journal of the United Nations, 2:33–41, 1984.Google Scholar
  16. 16.
    G. Sande. Confidentiality and polyhedra, an analysis of suppressed entries on cross tabulations. Technical report, Statistics Canada, unknown date.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1995

Authors and Affiliations

  • Ming-Yang Kao
    • 1
  1. 1.Department of Computer ScienceDuke UniversityDurhamUSA

Personalised recommendations