US 11,861,038 B2
Secure multiparty differentially private median computation
Jonas Boehler, Karlsruhe (DE); and Florian Kerschbaum, Waterloo (CA)
Assigned to SAP SE, Walldorf (DE)
Filed by SAP SE, Walldorf (DE)
Filed on Dec. 2, 2019, as Appl. No. 16/699,997.
Prior Publication US 2021/0165906 A1, Jun. 3, 2021
Int. Cl. G06F 21/00 (2013.01); G06F 21/62 (2013.01); G06F 7/08 (2006.01)
CPC G06F 21/6245 (2013.01) [G06F 7/08 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A system comprising:
at least one hardware processor; and
a computer-readable medium storing instructions that, when executed by the at least one hardware processor, cause the at least one hardware processor to perform operations comprising, at a first party in a multiparty system having a plurality of parties with their own independent databases:
receiving a request via a network to evaluate a function over a range of data contained in a database of the first party;
dividing the range into a plurality of subranges of data;
securely computing, for each of the plurality of subranges of data, a selection probability for the subrange, based on information received about each of the plurality of subranges of data from at least one other party in the multiparty system, the at least one other party in the multiparty system providing the information based on data not shared with the first party;
selecting one of the plurality of subranges of data based on the selection probability;
dividing the selected subrange into additional subranges;
recursively iterating the computing, the selecting, dividing, and repeating for the additional subranges, until the selected subrange has a size of one, each iteration being performed on subranges of the selected subrange from an immediately previous iteration; and
returning a result via the network to the request, the request including data from the selected subrange with a size of one.