DatashareNetwork: A Decentralized Privacy-Preserving Search Engine for Investigative Journalists
Authors:
Kasra EdalatNejad,
Wouter Lueks,
Julien Pierre Martin,
Soline Ledésert,
Anne L'Hôte,
Bruno Thomas,
Laurent Girod,
Carmela Troncoso
Abstract:
Investigative journalists collect large numbers of digital documents during their investigations. These documents can greatly benefit other journalists' work. However, many of these documents contain sensitive information. Hence, possessing such documents can endanger reporters, their stories, and their sources. Consequently, many documents are used only for single, local, investigations.
We pre…
▽ More
Investigative journalists collect large numbers of digital documents during their investigations. These documents can greatly benefit other journalists' work. However, many of these documents contain sensitive information. Hence, possessing such documents can endanger reporters, their stories, and their sources. Consequently, many documents are used only for single, local, investigations.
We present DatashareNetwork, a decentralized and privacy-preserving search system that enables journalists worldwide to find documents via a dedicated network of peers. DatashareNetwork combines well-known anonymous authentication mechanisms and anonymous communication primitives, a novel asynchronous messaging system, and a novel multi-set private set intersection protocol (MS-PSI) into a *decentralized peer-to-peer private document search engine*. We prove that DatashareNetwork is secure; and show using a prototype implementation that it scales to thousands of users and millions of documents.
△ Less
Submitted 30 July, 2020; v1 submitted 29 May, 2020;
originally announced May 2020.
