Projects
Dia-Pol
Polarization or dialogue? A deep learning study of the Black Lives Matter and Me Too online social movements. Using the cases of Black Lives Matter and Me Too, Dia-Pol studies the conditions under which social media interactions between social movements and their countermovements produce polarization, and the themes on which these interactions foster dialogue. The project applies topic modeling and deep learning to test social movement theories. It was funded by a Marie Skłodowska-Curie Individual Fellowship (H2020-MSCA-IF-2020, grant 101028566). Project results on CORDIS.
HeForShe
Bogazici University Scientific Research Project 18241.
Computational Social Science Lab
I direct the Computational Social Science Lab at Bogazici, a group of students and researchers who work with agent-based modeling, machine learning, and text mining. We meet weekly for hands-on sessions and collaborate on research projects on social questions. Meet the team.
Dissertation
Making autocrats accountable: Interests, priorities, and cooperation for regime change (University of Pennsylvania, 2016)
When do opposition groups unite to overthrow an authoritarian government? This study examines the conditions under which opposition groups, despite persistent ideological conflicts, achieve the long-term cooperation needed to transform the status quo. Conventional theories predict cooperation among opposition groups that share anti-regime interests. Yet opposition groups often remain divided by ideological conflict despite these shared interests, as in the Syrian opposition to the Assad government. Using case studies of the transition to constitutional monarchy in the Ottoman Empire (1876–1908) and France (1814–1830) combined with network analysis, the study shows that cooperation becomes more likely when groups harmonize their strategies of power transition and postpone ideological conflicts. Groups also coalesce around the actor that shows commitment to effective strategies.
The study develops the concept of de-prioritization, the process by which actors postpone the resolution of certain disagreements. Agreeing on ideology is hard, but de-prioritized ideological differences no longer obstruct cooperation. Once ideological conflicts are sidelined, pragmatic actors abandon ineffective strategies through learning, which facilitates coordination on a cooperative solution. As long as ideological conflicts persist, opposition groups remain divided and at best achieve temporary gains that are reversed during authoritarian retrenchment. The research was supported by the Penn SAS Teece Fellowship.