Inference of Three-Way Table Entries from Two-Dimensional Projections

Multi-dimensional tables of sensitive information are often summarized and made public by means of lower-dimensional projections, which are intended to prevent any disclosure of con dential data. Multiple projections of the same underlying table are linked over common attributes, however, so there is concern about the possibility of recovering sensitive data by combining projections. In this paper, we present an algorithm that gives tight upper and lower bounds on cell values of three-dimensional data when the three twodimensional projections of that data are available.