Effective moderation of polarization on social media

Determine practical and effective mechanisms by which social media platforms can moderate political polarization among users, specifying actionable approaches that reduce echo chambers and improve cross-ideological information flow on platforms such as Twitter/X without relying on ad hoc assumptions about user behavior or network structure.

Background

The dissertation documents strong political echo chambers within COVID-19 discourse on Twitter, particularly a dense right-leaning bubble where information rarely enters or exits. Influential users tend to be partisan, and random-walk analyses confirm asymmetric communication patterns. These findings underscore the challenge of reaching audiences across ideological divides during public health crises.

Given these entrenched dynamics, the authors note that deciding whether platforms should intervene is debated, but the operational question of how to moderate polarization—i.e., concrete strategies platforms could implement to mitigate echo chambers—remains unresolved. A rigorous, evidence-based specification of interventions is needed to move beyond high-level aspirations.