Our data science team, led by Trina Renyolds-Tyler, supported by the Human Rights Data Analysis Group (HRDAG) and shaped by the community, built a machine-learning tool based on volunteer analysis of police complaint records.
Over the course of five months, more than 200 total volunteers met weekly to read narrative text from the documents made public through the lawsuit Green v. CPD. From those documents, we identified 27,000 complaints filed against the Chicago Police Department between 2011 and 2015. Our volunteers labeled more than 4,000 complaints and tagged violations they observed to train our algorithm, named Judy.
Learn more about Beneath the Surface and Judy here.