This $400,000 Project Grant awarded by the National Science Foundation Division of Information and Intelligent Systems under CFDA 47.070 (Computer and Information Science and Engineering) supports collaborative research to develop responsible methods for the design and validation of algorithmic rankers. The key objectives are to:
Quantify the impact of item attributes and engineering choices on ranked outcomes to guide the selection of appropriate scoring functions.
Quantify the impact of data uncertainty and scoring formula changes on ranked outcomes to enable the design of stable rankers.
Quantify and address fairness issues in ranked outcomes for underrepresented or disadvantaged groups.
The research is focused on score-based algorithmic rankers used in critical domains like hiring, admissions, and college rankings. The project aims to enable technical interventions and identify cases where alternative approaches are needed. The award is held by the Regents of the University of Michigan and runs from September 1, 2023 to August 31, 2027.