Songpo Yang （杨松坡）

Greetings!

I am a Boya Postdoctoral Fellow at the School of International Studies, Peking University. At PKU, I work with Xun Pang on research at the intersection of social science and artificial intelligence, and I serve as Assistant Director of the PKU Analytics Lab for Global Risk Politics.

I am a comparative political scientist who studies political elites and the institutions that select and shape them across political systems. My substantive work spans elite politics, Chinese politics, and global governance. Across these settings, I ask who governs, how elites rise, and how institutional position shapes political behavior. I also develop the data infrastructure needed to answer these questions, including comparative political biographies, auditable AI-assisted workflows, and multimodal measures derived from images, voice, and video.

My work has appeared in the American Political Science Review and International Affairs. A related methodological project on agentic political-biography extraction has received a revise-and-resubmit invitation as a Research Note from the American Journal of Political Science.

I received my PhD in political science from the Department of International Relations, Tsinghua University and was a visiting PhD student in the Department of Political Science at Columbia University from September 2022 to October 2023. During my time at Columbia, I was advised by Junyan Jiang and Naoki Egami. I also hold an M.A. from Johns Hopkins SAIS, an M.L. from Tsinghua University, and a B.A. from the University of International Business and Economics. Before beginning my doctoral studies, I worked at CITIC Securities Asset Management.

Featured Research

“Agentic Framework for Political Biography Extraction.”
with Yifei Zhu, Jiangnan Zhu, and Junyan Jiang.
Revise and resubmit (Research Note), American Journal of Political Science.
This project develops and evaluates an auditable synthesis-and-coding framework for turning fragmented, multilingual sources into structured political biographies. It shows that data quality depends on the full evidence-production system, not only on the underlying language model.

View all research →    Explore data and methods →

Connect

For research, data-access, or collaboration inquiries, please contact me at yangsp@pku.edu.cn. You can also explore Risk-A-Lab Data Intelligence, view current Research Opportunities, or find code and project materials on GitHub.

          /\
         /**\
        /****\   /\
       /      \ /**\
      /  /\    /    \        /\    /\      /\
     /  /  \  /      \      /  \/\/  \    /  \/\                           -------- __@
    /  /    \/ /\     \    /    \ \   \  /   /  \       LIFE IS SHORT      ----- _`\<,_
   /  /      \/  \/\   \  /      \     \         \      KEEP FORWARD         ---- (*)/ (*)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~