Kaleigh Clary Named 2018 Data Science for Social Good Summer Fellow

Kaleigh Clary, a Ph.D. student in the University of Massachusetts Amherst's College of Information and Computer Sciences (CICS), has been named a 2018 Data Science for Social Good Summer Fellow.

Run by the University of Chicago, the summer-long fellowship program looks for students in the computer science, statistics and social sciences fields who are interested in data science along with making an impact on the world.

Aspiring data scientists attend the university summer program to work on data mining, machine learning, big data, and data science projects with social impact. Fellows' work includes collaborating with non-profits and government to solve issues in education, health, energy, public safety, transportation, international development, economic development, and more worldwide problems.

After getting her B.A. from Hendrix College in computer science and mathematics, Clary enrolled in the University of Massachusetts Amherst's computer science master's program, and is working toward earning her Ph.D. in 2020. Over the summer of 2017, Clary worked as an intern in the MIT Lincoln Laboratory. She previously interned as a data analyst for Acxiom in Arkansas.

She is currently a research assistant in the CICS Knowledge Discovery Laboratory where she is advised Professor David Jensen.

Clary's research focuses on causal inference and experimental design, specifically for applications in computational social science involving a relational component and those regarding fairness in machine learning.

"When we use causal models to describe, for example, systems of oppression, we are implicitly reasoning about the effects of interventions in that system that could perhaps dismantle or disrupt those systems," Clary wrote in her fellowship application. "I am proud of having identified a connection between issues I am deeply passionate about -- power structures, inequality, and intersectional feminism -- and my graduate work in causal reasoning."

Clary has noticed issues caused by bias in data that is unresolved during the modeling procedure. By bringing causal reasoning to data-generating processes, Clary hopes to address fairness, accountability, transparency, and ethics in machine learning.

Written by Jill Webb (Journalism '18)