Department of Computer Science
Stony Brook, NY 11794-2424
Robert F. Kelly received his Ph.D. in Computer Science from NYU Courant Institute of Mathematical Sciences in 1991.
Research deals with computational problems in political informatics, specifically systems for the analysis of issues related to congressional redistricting. Research is multidisciplinary, involving topics in computer science, applied mathematics, and political science. Student involvement in research is enabled through the Stony Brook University VIP initiative (PoliTech program) as well as graduate and undergraduate research in Computer Science. At the center of the research is the Automated Redistricting System (ARS), a system for the rapid generation of statewide congressional districts in accordance with constitutional and court-ordered guidelines, as well as user-defined preferences. ARS incorporates measures of the quality of a districting plan into a mathematical function, and then attempts to optimize it. Once the user selects a state, she can specify the importance of various measures such as compactness, equal population, and political fairness, along with other parameters such as the desired number of congressional districts. The system includes a database containing election district geographic data, election result data, and statewide demographic data for many of the states subject to redistricting following the 2020 Census. Individual research topics include:
- Algorithms for graph partitioning;
- Efficient algorithms for the measurement of geographic compactness;
- Measures of political fairness, including measures of gerrymandering;
- Techniques for the visualization of multi-parameter data including racial, economic, and voter preferences;
- Probabilistic analysis specific redistricting plans;
- Parallel software for the implementation of compactness, political fairness, and graph partitioning algorithms; and
- Legislation for redistricting requirements, voting rights, and election data quality.