Reputation Systems Against Social Engineering and Manipulation in Online Environments
Date: 17th Oct – ThursdayTime: 15:00-16:00Location: Ballroom A
The modern online social environments such as messaging systems, social networks and marketplaces provide high connectivity between participants and high interaction speeds. That makes it impossible for a participant to access all risks of social engineering and opinion manipulation that people are exposed to.
Most of known electronic communication environments, social networks, online media ecosystems, marketplaces and unmanaged transactional system such as blockchains easily host scam and manipulation campaigns targeting individual wallets, corporate assets and public security. Even the rating systems being employed in most of marketplaces are vulnerable for reputation gaming attacks with different strategies.
We discuss our proposition for flexible “liquid rank” reputation assessment system based on graph theory and “liquid rank” implemented as open source. We present how different configurations of the reputation system may reduce loss to scam for honest participants and increase costs of scam for dishonest ones, with examples for real blockchains and simulations of marketplaces. We also discuss possible design and implementation options for such system.
In the end, the weakest chain in computer security is a human, so the social engineering may crack what is not cracked exploiting software and hardware vulnerabilities. To address this, we present the notion of personal security assistant based on the cognitive model of the human mind and intended to defend human users from manipulation and negative impacts.