The study of police language aims to find patterns that can lead to tragic consequences
Adverse encounters between police officers and young men from under-represented backgrounds attract significant national attention to social justice issues and have been labeled a public health issue by several organizations. Now, thanks to a new four-year, $ 2.75 million grant from the National Institutes of Health, an interdisciplinary team of researchers aims to examine transcripts of police radio communications to observe what happens during these encounters and study patterns of interaction that can lead to unfortunate or tragic results.
“We hope to identify signals in the language, such as vocabulary and speech, which suggest that a meeting between a law enforcement officer and a young man belonging to a male minority is going to take a turn for the worse. worse, ”said Shomir Wilson, assistant professor in Penn State. College of Information Science and Technology. “Language conveys a lot of information about a person’s state of mind, actions, mood and comfort level.”
Working with human development experts at the University of Chicago, Wilson will lead a team at Penn State to use natural language processing to extract large-scale information from Chicago-area police scanner transcripts. His team will also carefully examine the privacy ramifications of police radio communications in general and the dataset in particular.
“Law enforcement officers frequently use their radios to report what they encounter, and they use a combination of standard jargon and open language to quickly describe situations,” Wilson said. “We want to go beyond literal descriptions and try to deduce what the police think and assume during encounters. If we can do this, it is a step towards identifying strategies that will defuse opposing encounters. “
The interdisciplinary project will combine research in natural language processing, computational social sciences and privacy. Penn State’s contribution will include developing automated methods to sort a large volume of transcripts, using supervised and unsupervised machine learning to explore transcripts, and studying the structure of incidents to be able to identify distinguishing features in language that can predict the outcome of incidents. The Penn State team will also identify potentially sensitive data and determine the best approach to share it with the research community while protecting the identity of those involved.
In addition to Wilson, core members of the research team include joint project leaders Chris Graziul and Margaret Beale Spencer, both at the University of Chicago; Karen Livescu of the Toyota Technological Institute in Chicago; and Lisa Thurau of Strategies for Youth, a nonprofit organization that works to improve interactions between police and young people. Pranav Venkit, a PhD student in computer science, is also part of the Penn State team.