Computing Innovation Fellow joins Tufts

Nasim Sonboli, a PhD candidate at the University of Colorado, Boulder, will join the Department of Computer Science as a Computing Innovation Fellow (CIFellow). She will work with Bert Huang, who is an assistant professor in the Department of Computer Science and the Data Intensive Studies Center (DISC). Sonboli’s expertise is in algorithmic fairness of recommender systems. People interact with these systems every day, and only recently have researchers begun to understand how to train these systems to treat their users more equitably. 

Sonboli and Huang will collaborate on a project investigating how algorithms can encourage fairness and deter discrimination, while also adhering to the principle of data minimization. Data minimization dictates that computing systems should only collect the minimal amount of data about people as necessary, thus reducing the potential for privacy breaches. Unfortunately, many algorithms that aim to be fair and equitable rely on knowing protected attributes of their users. 

For example, knowing the race of users allows systems to measure how they treat users who are Black as compared to users who are Asian. These systems can then use that information to balance themselves and provide equal service to people of different races. But this information may not be directly relevant for the type of service the system provides. Consider a scenario of asking search engine users for their race before they can look up movie reviews. Such information gathering often violates data minimization. Sonboli’s research seeks to answer the question of how to balance data minimization with fairness.

The CIFellows program provides fully funded postdoctoral fellowships for recent PhD graduates. The program is funded by the National Science Foundation (NSF) and is administered by the Computing Research Association (CRA) and the Computing Community Consortium (CCC). CIFellows are provided funding for two years as postdoctoral researchers and are chosen in a highly selective process that considers the fellows’ research accomplishments and the impact of their proposed research.