Social media and online social networks have had a profound impact on society, particularly in fast-changing situations where the right information needs to quickly reach specific people. Examples of such situations include political movements or revolutions, natural disasters, and financial crises.

In such disruptive environments, social networks and people’s behavior are dynamic — they react and adapt to the situation. This creates changes in the network structure of the platform, which in turn affects the dynamics of information diffusion. We do not fully understand how these network structures and behaviors operate around times of instability or their impact on information diffusion. This proposal aims to close this gap through four avenues of investigation:

1. Measuring and comparing network structure, user behaviors and content shared in social media before, during and after a diverse and important set of exogenous shocks, specifically: natural disasters, poll changes and unexpected outcomes during political elections, and financial market changes
2. Identifying the long-term effects of shocks on social media and the typical phases of community recovery
3. Predicting political and financial shocks from changes in network structure and user behavior
4. Characterizing the impact of changes in network structure, user behavior and content on the dynamics of information diffusion via modeling and simulations