Libreflix: plataforma de vídeo sob demanda peer-to-peer com recomendações que preservam a privacidade para um streaming livre

Abstract

This work aims to address two complementary fronts. The first, which constitutes the core of this research, presents Libreflix, an on-demand video streaming platform designed for the free publication and consumption of audiovisual content. Unlike traditional platforms, typically controlled by corporations that impose restrictive copyright policies and charge for access, Libreflix aggregates content under permissive licenses, fostering collaboration, community engagement, and the democratization of knowledge on the internet. The platform’s architecture and its main functionalities are described, along with relevant aspects and lessons learned throughout its ongoing development. The second front of this work explores the use of Trusted Execution Environments (TEE), with an emphasis on Intel SGX technology, for the management of recommendation systems in Online Social Networks – in this case, within Libreflix itself. This approach seeks to protect user data and prevent its misuse, without compromising the functionality or financial sustainability of these networks. The architecture of the proposed solution and performance analyses are presented, providing guidance for selecting recommendation algorithms suitable for secure execution within Trusted Execution Environments.