Functional dependencies and their connection with security of statistical databases |
| |
| Affiliation: | 1. From the Department of Ophthalmology, Duke University Medical Center Durham, North Carolina, USA;2. Department of Ophthalmology, University of California, San Francisco, California, USA;3. Jaeb Center for Health Research, Tampa, Florida, USA;4. Institut de la Vision, Sorbonne Université, INSERM, CNRS, Paris, France;5. Centre Hospitalier National d''Ophtalmologie des Quinze-Vingts, INSERM-DGOS CIC1423, Paris, France;6. Retina Foundation of the Southwest, Dallas, Texas, USA;7. Medical College of Wisconsin, Milwaukee, Wisconsin, USA;8. Foundation Fighting Blindness, Columbia, Maryland, USA;9. Kellogg Eye Center, University of Michigan, Ann Arbor, Michigan, USA;10. Department of Biomedical Engineering, Duke University, Durham, North Carolina, USA;11. University of Toronto and Hospital for Sick Children, Toronto, Ontario, Canada;12. National Eye Institute, Bethesda, Maryland, USA;13. Vitreoretinal Associates, Gainesville, Florida, USA;14. Moorfields Eye Hospital and UCL Institute of Ophthalmology (M.M.), London, United Kingdom;15. Casey Eye Institute, Oregon Health & Science University, Portland, Oregon, USA;p. Baylor College of Medicine, Houston, Texas;1. University of Louisville, Louisville, Kentucky, USA;2. University of Denver, Denver, Colorado, USA |
| |
| Abstract: | The term “security of a statistical database” usually refers to the problem of inference control. There are two main categories of inference controls in statistical databases: restrictions and perturbations. The influence of functional dependencies (which may hold among attributes in a statistical database) on restriction techniques is studied. For some techniques, either on the table or cell level, we demonstrate how these dependencies might lead to a compromise. It is shown that functional dependencies may decrease the effectiveness of several known inference controls (more precisely: output restriction controls). For some methods we give the way to overcome nondesirable influence of dependencies on control techniques. |
| |
| Keywords: | |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
