Xavier Students Measure Coronavirus Fear Via Twitter
May 20, 2020
Social media has become the default source for information about COVID-19, making it the ideal place to study how people feel about the virus.
To better understand how people are reacting to the pandemic, students in Yemi Oyediran’s Software Engineering course created a data-mining application to track sentiment about COVID-19 in the United States. As part of their software engineering course project, these students designed a website to gather information from Twitter. The application launched to the public in mid-May.
“I've been blown away by the initiative and the level of commitment these students have displayed all semester. This project and these students are just a partial reflection of the level of commitment that Xavier University brings to our community and our country,” Oyediran said.
Just before spring break, Oyediran, adjunct professor of Computer Science, assigned the project, which was intended initially to focus on studying political data on Twitter. As things rapidly shifted during the course of the break because of the pandemic, the project evolved to focus on a more timely topic: COVID-19. This change, driven by the students, created an opportunity for them not only to build on their skills as computer scientists, but also to use their knowledge of social media technologies for the common good.
“This is definitely the most amount of hours we've put in on an assignment, but it's for something we all believe in," said Dustin Sapp, a student in the course.
Data-mining is a process in which large databases, like Twitter, are examined in order to generate new information. Students in Oyediran’s course are using this technique to access a constant stream of tweets expressing sentiment about COVID-19. They collect any tweets, hashtags and even emojis that relate to COVID-19. Once collected, the tweets are ranked based on a number of criteria.
Students leverage IBM Watson machine learning tools to track tone and emotion in the tweets, plus Microsoft Adder’s Tools to determine levels of positivity and negativity. During the first 15 minutes after the application was launched, more than 7,000 tweets had been analyzed. The number continues to grow.
Beyond showcasing impressive computing skills, students working on this project embraced the heart of their liberal arts education. Pulling from philosophy, theology and history courses required as part of the Core Curriculum, students examined the ethical responsibility involved in data collection. They dedicated multiple hours of out-of-class time to discuss the ethical concerns that may be associated with studying this data.
“No one wanted to leave, even though class was over, until they fully discussed every aspect of the ethical concerns and determined what the ethical boundaries for data collection should be,” Oyediran said. “It’s a Xavier education in full effect."
Reflecting on why the project has been so successful, Oyediran went on to explain how this project gives students an environment where they can find solutions to real-world problems.
“I'm very excited to see the long-term impact that this experience will have on their careers as technologists,” he said. “I’m sure each of them will make an impact on their communities. These are the type of opportunities we seek to provide at Xavier University.”
By Jade Martinez, College of Arts and Sciences
Feature Image: Students in Yemi Oyediran’s class meet remotely to work on their project.