UC Berkeley researchers win $2M grant for criminal justice datasets

from left to right, top to bottom: Aditya Parameswaran, Niloufar Salehi, Sarah Chasins, Erin Kerrison, Joseph Hellerstein
From left to right, top to bottom: Aditya Parameswaran, Niloufar Salehi, Sarah Chasins, Erin Kerrison and Joseph Hellerstein won a grant to create criminal justice datasets that will help defense attorneys.

Related Posts

Researchers at UC Berkeley received a $2 million grant from the National Science Foundation, or NSF, for the creation of criminal justice datasets aiming to assist defense attorneys.

The project aims to create an accessible “data portal” containing police misconduct information for defense attorneys to reference on behalf of their clients, according to Aditya Parameswaran, assistant professor at the UC Berkeley School of Information and in electrical engineering and computer sciences, or EECS.

“Public defenders often have barely a few hours to prepare their case and are juggling many cases at once, so any tooling that will help them quickly paint a convincing picture for the jury, backed by data, is invaluable,” Parameswaran said in an email. “Some early work by the Legal Aid Society in NYC showed that manually gathering a collection of police misconduct data led to a number of cases against innocent clients that were dismissed.”

The project’s primary authors included Parameswaran; Joe Hellerstein, campus computer science professor; Sarah Chasins, campus assistant professor of EECS; Niloufar Salehi, assistant professor at the School of Information; and Erin Kerrison, campus assistant professor at the School of Social Welfare, Parameswaran noted.

According to Dan Cosley, a program director in the NSF Information and Intelligent Systems division, NSF will provide approximately $700,000 per year over three years to fund the project’s execution.

“What set this proposal apart is that the justice system is an important work context,” Cosley said in an email. “Better tools for analyzing legal data could impact the quality of the U.S. justice system.”

Chasins noted that the funds will go toward graduate students developing technological solutions for organizing vast datasets of police misconduct into a digestible format.

She added that this data will not only assist public defenders but also journalists and other individuals working against police misconduct.

“One outcome that’s particularly inspiring to me is the possibility of sparking systemic change,” Chasins said in an email. “Some of our collaborators at the National Association of Criminal Defense Lawyers have used data like this to pressure individual cities to repeal police secrecy laws–laws that keep records of police misconduct hidden from the public.”

Parameswaran noted that as public defenders remain underresourced, this project aims to equip them with unique data tools.

The researcher’s “secondary goal” is to create general resources for data organization that do not require coding expertise, Parameswaran added.

“The hope would be that these tools will not just apply for public defender needs across various jurisdictions in CA to start, but also across the US, and more broadly, even going beyond the criminal justice domain to other similarly under-resourced domains,” Parameswaran said in an email.

Contact Rachel Raps at [email protected], and follow her on Twitter at @rachelraps_dc.