Illustrated profile photo

Hey, I'm Jeffrey (he/him/他)! I study computer science and politics, with a concentration in comparative media studies, at MIT. I'm interested in using technology to bring clarity and transparency to politics, and have done work around data journalism, digital censorship/propaganda, surveillance, policing, and political tech.

Right now, I'm researching Amazon Ring's growing surveillance network of smart doorbell cameras in the MIT Media Lab's Human Dynamics Group. I also am a 2020-2021 Human Rights & Technology Fellow at the MIT Center for International Studies, where I study privacy and policing concerns related to acoustic gunshot detection systems.

Please reach out via email or Twitter!

Portfolio

Research

Screenshot of JSON metadata

Fall 2018

Russian Internet Disinformation

Used machine learning and sentiment analysis to explore Russian disinformation on Twitter in the 2016 and 2018 elections. Wrote an interactive data journalism series and created a (now defunct) Twitter bot. Completed as part of an Independent Project at Phillips Academy with R. Miles McCain, under the mentorship of John Palfrey, Michael Barker, and Malgorzata Stergios.

Tiananmen Square with security cameras in the foreground

Fall 2018

Deciphering Chinese Censorship

Developed the Weibo Censorship Dashboard, a nearly real-time, fully open-source feed of censorship on the popular Chinese social media site Weibo. Published paper on SSRN. Completed as part of an Independent Project at Phillips Academy with R. Miles McCain, under the mentorship of John Palfrey, Michael Barker, and Malgorzata Stergios.

Data Visualization

Green choropleth map of Los Angeles

Spring 2019 - Present

Blog

A data visualization publication covering politics, society, and culture. Example topics include mapping the inequal access to shade in Greater Los Angeles and quantifying how lyrical modern hip-hop is.

Screenshot of website, with a linechart and numbers

Summer 2020

Amazon Ring's Police Partnership Tracker

Created a website monitoring the over 1,400 partnerships Amazon Ring has with local law enforcement. Built with Node.js, D3, and MongoDB.

Samuel Phillips Hall at Phillips Academy, a Greek-style building

Spring 2018

State of the Academy

Created a website of interactive visualizations showing the results of The Phillipian's State of the Academy survey, an annual survey covering politics, academics, wellness, and culture at my high school. Built with D3.

Politics

Blue choropleth map of Texas's 22nd Congressional District (near Houston)

Jan 2020 - Nov 2020

Bluebonnet Data Fellow

Volunteer data analyst, used Python and GIS to help down-ballot, Democratic candidates win in swing districts.

Screenshot of the homepage of LobbyView

Jan 2020 - Aug 2020

LobbyView

Worked with the MIT Political Science Department, developed interactive data visualizations using D3 and React for LobbyView, a database of Congressional lobbying that aims to facilitate political science research and bring transparency to money in politics.

Screenshot of a gerrymandered district near Baltimore, MD

2018

Is My District Gerrymandered?

Created an interactive website that estimates how gerrymandered a U.S. congressional district is, using geometry and election results. Project won the 2018 Congressional App Challenge in Massachusetts's 6th District.

Screenshot of website, showing a line chart

Spring 2018

Exploring the Seats-Votes Curve

Developed an interactive, data-driven explainer of the seats-votes curve, a measure of fairness in redistricting. Built with D3.

Full Stack Development

Red choropleth map of the world's countries

Summer 2020

The Power of Without

Worked with the City Science Group, in the MIT Media Lab, to develop a crowdsourced data collection and visualization platform for informal settlements in developing countries. Created with Node.js and Leaflet.

Screenshot of website

Spring 2019

PollPA

Developed an open source, anonymous platform to poll students at my high school on important issues, in real-time. Created with R. Miles McCain, using Django and D3.