Discovery Exchange Students Analyze Tweets to Map Pandemic Reactions in COVID-Twitter Project

COVID Twitter project
August 13, 2020

The COVID-Twitter Project, created by Ameek Bindra (Berkeley ‘22) and a team of undergraduates across the country, analyzes tweets about COVID-19 to map how reactions to the pandemic changed over the months. According to Ameek, the Discovery Exchange provided invaluable resources and guidance that helped the project come to fruition. (Note: responses modified for length) Discovery Exchange is a new initiative that originated from the Discovery Program as an online hub for Berkeley students doing independent data science projects. The Discovery Program connects undergraduates with hands-on, team-based opportunities to contribute to cutting-edge data research projects. 

Question: What is your data science research about? 

With the re-opening of businesses across the country, there are fears that the number of COVID-19 cases will rise again. How is the curve affected by how people respond to COVID-19? 

Answer: We used machine learning and Twitter data to map people’s attitudes towards the pandemic and see how that has changed over the past few months. In addition, we analyzed the demographics of users who tweet about COVID-19 as well as over 150k tweets across 62 languages and hundreds of countries. We plan to display our findings on Dash, a web app that supports interactive data visualizations. The goal is to use this data and sentiment analysis to identify geospatial locations that will need the most resources to get back and running after the pandemic.


Q: How has being a part of the Discovery Exchange helped you continue with your research under this remote work environment?

A: Discovery Exchange has been able to provide my team and me with invaluable resources when conducting this project. In fact, without Exchange, I would not have begun this project in the first place. It was on the Piazza where I found my initial team and had the momentum to carry on with this project. Whenever I felt stuck, I got a lot of valuable advice from Arlo on what to do next. As summer classes kicked in, a lot of team members were not able to dedicate as much time anymore, so having Arlo and the Exchange program consistently cheer us on was a huge help to making the project come to fruition.


Teamwork taps into different skill sets, institutions

“The most exciting part of our project was working with Berkeley students with very different academic backgrounds and skillsets, along with students from other institutions across the country like Carnegie Mellon University (CMU), University of Florida, University of Michigan, and the University of Illinois at Urbana-Champaign (UIUC),” Bindra said. “Virtually collaborating with a huge team has been a great experience, especially with the heavy exchange of ideas and discourse on how to further analyze data.”

COVID Twitter project team


Pictured: (from left to right, then by row): Vanita Sharma (University of Michigan '22), Ameek Bindra (Berkeley '22), Tanya Krishan (Berkeley '22), Neha Haq (Berkeley '22), Selena Lu (Berkeley '22), Jing Low (University of Florida '22), Jane Zhang (Berkeley '22), Aditi Tyagi (UIUC '22), Kimberly Lo (CMU '21).

Not pictured: Derek Cai (Berkeley '21), Shruthi Shridhar (UW-Madison, MS)