The talk will focus on how Facebook develops, implements, and enforces its policies on what content and behavior is not allowed on the platform and how research informs these processes. This will include details on the internal policy development processes, how product and operations teams work to build the machine and human review processes, as well as details on the different ways in which both theoretic and empirical social science research inform these efforts. The talk will conclude with Facebook’s approach to collaboration with external researchers.

Recommendation systems are ubiquitous and impact many domains; they have the potential to influence product consumption, individuals’ perceptions of the world, and life-altering decisions. These systems are often evaluated or trained with data from users already exposed to algorithmic recommendations; this creates a pernicious feedback loop. We demonstrate how using data confounded in this way homogenizes user behavior without increasing utility.