Data science approaches for measuring the effects of foreign information operations

Authoritarian governments such as Russia, China, or Iran invest significant resources in their propaganda efforts both domestically and abroad. Externally, these regimes operate wide ecosystems of state-controlled media outlets in dozens of languages around the globe, and their diplomats have joined Western social media platform where they target foreign audiences with pro-regime propaganda and disinformation. These overt attempts to shift public opinion through overt public diplomacy are complemented by covert forms of exerting sharp power, including operating grey news outlets with hidden state ties as well as running coordinated networks of fake accounts in an effort to inauthentically manipulate social media platforms.

While the presence of such overt and covert efforts has been universally established, the academic literature still lacks a sufficient understanding of the effectiveness of foreign information operations targeting liberal democracies. This includes both the question of reach, thus how many individuals around the world are being exposed to foreign authoritarian propaganda, as well as its impact, in other words whether the content is actually able win hearts and minds by shifting individuals' attitudes and behaviours.

To close this gap, my DPhil research applies a mix of data science methodologies, including causal inference, natural language processing, network analysis and factorial experiments, to study the impact of authoritarian information operations on foreign audiences in liberal democracies. The results of my research will provide timeline insights and help policy makers and national security officials make informed choices in the delicate balancing act between upholding the freedoms of speech and expression that form the basis of liberal democracies on the one hand, and protecting these democracies from malign foreign interference on the other.