Privacy-preserving Information Sharing

10 September 2019 - 10 a.m.
Versione stampabile

Luogo: Room A222 Povo1, via Sommarive, 9 - Povo 
Ore: 10:00

Relatore:

  • Carlo Blundo (Università degli Studi di Salerno)

Abstract:

In this talk, we address privacy issues related to sharing information in a distributed system consisting of autonomous entities, each of which holds a private dataset. The entities are willing to share information computed as a function of their datasets. Nevertheless, no entity is willing to disclose its private data to other entities due to privacy concern. In this talk, we will focus on Private Set Intersection and its variants. Private Set Intersection (PSI) refers to a setting where two parties each hold a set of private items, and wish to learn the intersection of their sets without revealing any information except for the intersection itself. We will see how PSI can be used to solve several practical matching problems such as: find out whether any suspect is on a given flight, discover if tax evaders have accounts at foreign banks, execute a genetic paternity test, suggest advertisement recommendation, check document/multimedia similarity, run a biometric authentication protocol, steer shoppers within a shopping mall towards the shops selling their desired items without knowing the shopping list's content.

Referente: Massimiliano Sala