Authors:
Nouf Al-Juaid
1
;
2
;
Alexei Lisitsa
1
and
Sven Schewe
1
Affiliations:
1
Department of Computer Science, University of Liverpool, Liverpool, U.K.
;
2
Department of Information Technology, College of Computers and Information Technology, Taif University, Saudi Arabia
Keyword(s):
Graph Databases, Secure Multi-Party Computation, Federated Databases, Secure Data Processing.
Abstract:
We present a methodology for secure joint querying over federated graph databases based on secure multiparty computation (SMPC). Using SMPC instead of (or in addition to) encryption lifts reliance on the security of the encryption mechanism. The secret keeping is, instead, guaranteed by an SMPC protocol that protects the information required to answer a given query so that it is not shared in full on any communication line. We have recently outlined how this could be done in principle in a position paper, albeit with a sluggish implementation with an enormous computational overhead that rendered it unusable in practice. In this paper, we proposed an approach by better integrating it with the SMPC protocol, implementing it in JIFF, and covering the joint functionalities and languages of Conclave, Neo4j Fabric, and APOC. When implementing our prototype, we demonstrate how small queries can be served in fractions of a second, thus improving the performance of secure joint querying by tw
o orders of magnitude compared to the implementation in previous work while also significantly extending its set of supported queries.
(More)