Resume

Chang Ge

+1 734-882-0659 | changgge@umich.edu

PUBLICATIONS

1. Welsh, M., Rey, C.F., Ge, C., Nowak, T., Tomkins, S. Algorithms in the Stacks: Investigating data-driven, for-profit diversity audits in public libraries. Proceedings of the 2025 ACM Conference on Fairness, Accountability, and Transparency. Association for Computing Machinery, New York, NY, USA. [https://dl.acm.org/doi/full/10.1145/3715275.3732140]

   • Acceptance rate: 26.8%

2. Ge, C., Zhang, J., Xu, H., Krupnikov, Y., Bednar, J., Tomkins, S. What does the public want their local government to hear? A data-driven case study of public comments across the state of Michigan - Submitted to Journal of Quantitative Description, also available on arXiv [https://arxiv.org/pdf/2507.18431]

   • Implemented a supervised learning framework using RoBERTa and DistilBERT models to categorize public comments at city council meetings from 15 cities in Michigan in 2023

3. Ge, C., Gautam, N.,  Vajjhala, S., Ng, C., Hasell, A., Tomkins, S., Algorithmically Moderated Content Depending on News Appetite on TikTok: A Case Study around the 2024 US Election. - ICWSM '26 Revise and Resubmit

   • Engineered a 200-agent experimental platform to audit TikTok’s recommendation system, generating millions of interaction records for time-series causal analysis.

   • Built an automated AWS pipeline integrating Python and web automation, reducing manual setup time by 99% and enabling reproducible experiments at scale.

   • Developed a cascading RoBERTa classifier with a custom bias-aware metric, achieving a 15% performance improvement over GPT-4o, DistilBERT, and BART.

   • Delivered actionable insights on algorithmic bias and content exposure patterns, advancing transparency in recommendation systems.

4. Rothschild, D., Liu A., Thompson, A., Ge, C., Hasell, A., Tomkins, S., The descriptive role of identity and issues in partisanship and what we think elites know about it from TikTok. - Submitted to Journal of Quantitative Description

5. Abbadi, M., Chen Y., Ge, C., Lin, A.C., Toyama, K. (Alphabetical order) Value-Consistency Interventions Increase Trust in Muslims - Submitting to PNAS Nexus

   • Conducted a large field experiment in Michigan and an online experiment to compare the effects of two conversation-based interventions -- value consistency and perspective taking -- on American adults with respect to discrimination against Muslims

WORKING PAPERS

1. Dimant, E., Ge, C., Krupka, E., Robbett, A. (Alphabetical order) Escalation and Polarization Spillovers.

   • Conducted an online experiment coded in oTree on Prolific to explore how political identities affect people's reciprocity behaviors and spillovers. Tested (1) whether there is an initial interaction difference when participants are informed of their matched person's political identity; (2) whether there is a presence of escalation or spillover; and (3) whether there is an impact of a partisan identity on escalation or spillover.

2. Chen, Y., Ge, C., Li, L., Miller, D., Niederle, M. (Alphabetical order) Do Experimentally Elicited Preferences Predict Contributions to Wikipedia?

   • Built supervised Logistic Regression, Random Forest Regression, and Gradient Boosting Regression to predict if an individual would contribute to and how they would contribute to Wikipedia based on their cognitive abilities and social preferences.

3. Chen, Y., Ge, C., Li, L., Miller, D., Niederle, M. (Alphabetical order) Gender Difference in Contributions to Wikipedia.

   • Conducted a 3-year field experiment to examine the reason behind the phenomenon that less than 10% of Wikipedia editors are women

WORK IN PROGRESS

1. Nationalization indicated in local government meetings (with Sabina Tomkins, Yanna Krupnikov, and Jenna Bednar)

2. An inefficient two-party system and welfare loss (with David Rothschild, Sabina Tomkins, and Mirza Nayeem Ahmed )

3. Observatory of Attitudes Towards Public-Serving Institutions (OATLI) (with Sabina Tomkins and Ariel Hasell)

4. Book bans and intellectual freedom (with Sabina Tomkins and Madhumita Lahiri)

Conference Presentations

Oral Sessions

1. North American ESA Meeting @ Ohio State University, 2024

2. North American ESA Meeting @ University North Carolina Charlotte, 2023

Poster Sessions

1. Conference on Language Modeling (COLM) Workshop NLP4Democracy @ Montreal, Canada, 2025

2. Democracy's Information Dilemma @ University of Michigan, 2025

3. 2024 Norms and Behavioral Change (NoBeC) conference @ University of Pennsylvania, 2024

4. Midwest Speech and Language Days (MSLD) @ University of Michigan, 2024

5. Directions of Polarization, Social Norms, and Trust in Societies Perspectives from Behavioral Sciences @ MIT, 2023

REVIEWER SERVICE

Program Committee, Association for the Advancement of Artificial Intelligence (AAAI)-Artificial Intelligence for Social Impact Track (AISI)        2026

Association for the Advancement of Artificial Intelligence (AAAI)-Artificial Intelligence for Social Impact Track (AISI) 2025

Women in Machine Learning (WiML) @ Conference on Neural Information Processing Systems (NeurIPS) 2024

RESEARCH EXPERIENCE

• School of Information, University of Michigan, Ann Arbor, MI.       Nov. 2020 - Present

  • • Advised by Professor Sabina Tomkins, Justine Zhang, Erin Krupka, and Yan Chen

• Ross School of Business, University of Michigan, Ann Arbor, MI.   Jun. 2020 - Jun. 2021

  • • Advised by Professor Yue Maggie Zhou

• University of Michigan Medical  School, University of Michigan, Ann Arbor, MI.     Sep. 2019 - Dec. 2019

  • • Advised by Amy Huang, M.D., Director for Asia Programs

• National School of Development, Peking University, Beijing, China.     Aug. 2019 - Sep. 2019

  • • Advised by Shilin Zheng, Ph.D., Associate Researcher

• School of Business, Shanghai University of Finance and Economics, Shanghai, China.     Dec. 2017 - Mar. 2018

  • • Advised by Professor Heng Ju

TEACHING EXPERIENCE

Graduate Student Instructor Aug. 2022 - present

School of Information,  University of Michigan

1. SI 301: Models of Social Information Processing

2. CMPLXSYS/SOC 251: Computational Social Sciences

3. SI 670: Applied Machine Learning

Instructional Aid     Jun. 2021 - Jul. 2021

School of Information,  University of Michigan

1. SIADS 630: Causal Inference

Teaching Assistant   Jul. 2020 - May. 2021

Ross School of Business, University of Michigan

1. Strategy 503: Competing in the Global Business Environment (MBA core course)

2. Weekend MBA 512: The World Economy (Weekend MBA core course)

3. EMBA 639 - Global Business Environment

4. Fin 608: Capital Market and Investment Strategies (Master of Science in QFRM elective course)

Tutor Dec. 2019 - Apr. 2020

College of Literature, Science and the Arts, University of Michigan

1. Math 116: Calculus II

2. Math 215: Multivariable and Vector Calculus

3. Math 217: Linear Algebra

4. Statistics 412: Introduction to Probability and Statistics

INDUSTRY EXPERIENCE

A Better Community Consulting, Project Manager Intern   Beijing, Sep. 2019 - Jun. 2019

Euro-Asia Consulting, Business Analyst Intern     Shanghai, May. 2018 - Feb. 2019

Guotai Junan Securities Limited, Financial Analyst Intern     Shanghai, Dec. 2017 - Mar. 2018

Meritco Services, Consultant Intern     Shanghai, Aug. 2017 - Oct. 2017

Language Partners International Limited, Co-founder & Product Manager Hong Kong, Sep. 2015 - Dec. 2016

HONORS

1. Poster Presentation Award at Democracy's Information Dilemma Conference (USD 500) 2025

2. Rackham International Student Fellowship (1-2 students per department per year), 2022-2023 academic year (USD 10,000) 2022

3. Department of Economics student spotlight  (5 students per year), 2020-2021 academic year 2020

4. Summer Seminar Distinction Certificate issued by Chinese Academy of Social Sciences and Industrial Organization Review (Rank 6/100, CNY 2,000) 2019

5. National Graduation Scholarship – First Place (Rank 1/350, CNY 8,000) 2018

6. Interdisciplinary Contest in Modeling, The Consortium for Mathematics and Its Applications (COMAP) – Meritorious Winner (top 4% among global candidates) 2016

7. “Challenge Cup” National College Student Business Plan Competition – Silver Prize 2016

8. China National College Student “Innovation, Originality and Entrepreneurship” Challenge – 2nd Place 2016 

TECHNICAL STRENGTHS

Professional Exams: 

1. passed Chartered Financial Analyst (CFA) level 1 with top 10% score among global candidates

2. passed the Microeconomic Theory PhD field prelim at Department of Economics, University of Michigan (subjects: preference and choice theory, general equilibrium, game theory, and mechanism design)

Languages: Chinese (native), English (proficient), Japanese (proficient), French (intermediate)

Programming Languages: Python, R, JavaScript, SQL, Shell, LaTeX

Statistical & Research Software: STATA

Frameworks & Libraries: TensorFlow, PyTorch, Scikit-Learn, Huggingface, Ray, JAX, Keras, spaCy, NLTK, pandas, NumPy, Polars, Matplotlib, Seaborn, Plotly

Web Development & APIs: Flask, Django

Tools & Platforms: AWS, Heroku, oTree