Research Program

I have a life-time commitment to studying technology and democracy. Growing up in Taiwan, polarization and identity were regular topics of conversation at the dinner table. My background inspired my first computational research program: understanding how algorithms restructure online networks of deliberation in modern democracies. Past projects include profiling bots that spreading COVID-19 and election misinformation, and the role Instagram played in the George Floyd Protests. This program has grown to applying machine learning and social network analysis to different domains, including offshore finance and mentorship networks. I also helped invent an award-winning instrument.

My work has been featured in The Washington Post, the New York Times, Scientific American, and the Swiss National TV, and I collaborate with a few companies to better understand human behavior. For a full list of publications, please see my Google Scholar.

Key themes: Inequality on social networks (racial, financial, and gender), misinformation, social algorithms

Highlighted Research

Mathematical Prestige: Elitism and Inequality

Chang & Fu

In mathematics, who you know matters. A network of 200,000 advisors and students show concerted efforts by academic committees, such as prize giving, can either reinforce the existing elite or reshape its definition.

Complex Systems of Secrecy: Offshore System of Oligarchs

Chang, Harrington, Fu, & Rockmore

Following the invasion of Ukraine, sanctions on oligarchs came under fire due to the oligarchs’ successful evasion. We analyze the role of an overlooked but highly influential group: the secretive professional intermediaries.

How Instagram Facilitated the 2020 BLM Protests

Chang, Richardson, & Ferrara

During the Floyd protests, positive framing emerged from unlikely opinion leaders: entertainment and meme accounts. Through 1.1 million photos, we show how content creators and citizen journalists shaped one of America's largest human rights movements.

Liberals Engage with more Diverse and Toxic Content

Chang, Druckman, Ferrara, & Willer

Most Americans receive some political information through social media. Analyzing ten years of Twitter data (n=4 billion), we explore how liberals and conservatives engage with political information from elites.

Published Research

Writing and Op-eds

H. Chang, S. Ahn, M. Brichta, and J. de Vight (2023). “Will She Win? Gender, Electability, and the Presidential Primaries.” Journalism and Mass Communication Quarterly.

Writing and Op-eds

J. Tang, B. Yan, H. Chang, Y. Nan, L. Zhen, and A. Yang (2023). “Policy Communication in Times of Public Health Crises: Longitudinal Network Modeling of US Politician-Health Agency Interactions during the COVID-19 Pandemic. Computers in Human Behavior.

Writing and Op-eds

K. Felkner, H. Chang, E. Jang, and J. May (2023). "WinoQueer: A Community-in-the-Loop Benchmark for Anti-LGBTQ+ Bias in Large Language Models." Annual Meeting of the Association of Computational Lingusitics (ACL).

Writing and Op-eds

J. Jiang, E. Chen, L. Luceri, G. Muric, F. Pierri, H. Chang, and E. Ferrara (2023). "What are your Pronouns? Examining Gender Pronoun Usage on Twitter." International AAAI Conference on Web and Social Media (ICWSM)