The Michigan Institute for Data Science at the University of Michigan held the first day of its Data Science Annual Symposium Tuesday. The virtual symposium, which will continue on Wednesday, featured a keynote talk on feminism in the data science field and two research talks.

The symposium’s programming began with keynote speakers Lauren Klein, English professor at Emory University, and Catherine D’Ignazio, urban science professor at the Massachusetts Institute of Technology. The pair discussed their book, “Data Feminism,” which was published last February.

Klein and D’Ignazio opened by introducing what they said are the seven principles of data feminism: examining power; challenging power; rethinking binaries and hierarchies; elevating emotion and embodiment; embracing pluralism; considering context and making labor visible.

“We see data feminism as sort of part of this growing body of work that is holding corporate and government actors accountable for their sexist, racist and classist data products,” Klein said. “Hiring algorithms that demote applicants that went to all women’s schools, search algorithms that circulate negative stereotypes about Black girls — I mean the list could go on, and it does go on.”

Klein and D’Ignazio said they were motivated to write their book after learning about potentially problematic data science systems such as face detection and hiring algorithms. Though these systems are dangerous, they are still widely used, they said.

“In many ways what inspired us to write this book is this wave of pushback that has been happening all along really, but that has been amplified in the past five years in particular,” D’Ignazio said. “In fact, data isn’t particularly doing anything for social good. What it, in fact, is doing when we look at it in aggregate is really describing the same old oppression.”

The symposium then moved to research talks, which presented findings from current research projects at MIDAS.

Bhramar Mukherjee, professor and chair of biostatistics and Lauren Beesley, postdoctoral fellow in the School of Public Health gave the first talk, titled “The Testing Paradox for COVID-19.” Mukherjee said their research focuses on the relationship between increased COVID-19 testing and rising case counts.

“This case counter is completely meaningless unless you know something about the testing strategy, and even then it gets complicated, so we are going to share some snapshots of that complication in terms of a statistical model today,” Mukherjee said. “Our aim is to come up with a mathematical framework about how to allocate tests optimally across populations defined by symptom severity.”

Quan Nguyen, research fellow in the School of Information, discussed how data on student activity collected from nearly 11,000 Wi-Fi access points around campus can be used for campus planning. 

In his talk, “Students’ Mobility Patterns on Campus and the Implications for the Recovery of Campus Activities Post-Pandemic,” Nguyen said this data will help the University plan for future semesters. 

“There’s a lot of decisions that need to be considered in terms of campus planning during COVID-19, such as which courses to be delivered online or in-person, which group of students are more exposed, how do students actually interact with one another on campus,” Nguyen said. “From this data, we can infer information about how students move around campus.”

Daily News Contributor Isabelle Regent can be reached at iregent@umich.edu.

The COVID-19 pandemic has thrown challenges at all of us — including The Michigan Daily — but that hasn’t stopped our staff. We’re committed to reporting on the issues that matter most to the community where we live, learn and work. Your donations keep our journalism free and independent. You can support our work here.

For a weekly roundup of the best stories from The Michigan Daily, sign up for our newsletter here.