February 21, Computational Social Science Lab & NetSCI Lab Talk: Deen Freelon on using AI to study racism in the news

SC&I Logo

The Computational Social Science Lab Speaker Series will be co-hosted by the NetSCI Lab and will feature Presidential Professor Deen Freelon, University of Pennsylvania, who will offer insights into using AI tools to analyze media discourse about racism.

Deen Freelon
Analyzing frames around “racism” in the news media: An AI-assisted approach
Mediated discourse around racism has been a topic of occasional interest in political communication research over the past few decades, but traditional approaches have left vital questions unanswered. Much of this work is qualitative, offering penetrating theoretical insights but little in the way of generalizable results. The quantitative side of the literature is dominated by content analyses—both manual and automated--some of which merge racism with other offensive communications under the umbrella concept of “hate speech.” The current approach focuses specifically on the discursive contexts in which the term “raci(al/sm/st)” (including euphemisms like racial/ly ”charged,” “tinged,” “motivated,” “connotations,” etc.) appears in historical and contemporary mainstream, left-wing, right-wing, and Black news media. In keeping with definitions of framing that focus on problem definition and assignment of blame, this presentation will demonstrate the use of AI (specifically LLMs) to identify claimants, objects, perpetrators, and targets of racism and related concepts in mediated discourse. This project, which is in its initial planning phase, has two main goals: first, to identify the most important predictors of certain groups (e.g., racial, gender, age) appearing as perpetrators and targets of racism; and second, to produce a comprehensive, quantitative impression of how these trends have changed over time. 

Deen Freelon is a Presidential Professor at the Annenberg School for Communication. A widely recognized expert on digital politics and computational social science, he has authored or coauthored over 60 book chapters, funded reports, and articles in journals such as Nature, Science, and the Proceedings of the National Academy of Sciences. He was one of the first communication researchers to apply computational methods to social media data and has developed eight open-source research software packages. The first of these, ReCal, is a free online intercoder reliability service that has been running continuously since 2008 (when he was a Ph.D. student) and used by tens of thousands of researchers worldwide. He has been awarded over $6 million in research funding from grantmakers including the Knight Foundation, the Hewlett Foundation, the Spencer Foundation, and the US Institute for Peace. He was a founding member and remains Senior Researcher at the Center for Information, Technology, and Public Life at the University of North Carolina at Chapel Hill, one of five academic research centers in the Knight Research Network (est. 2019) to receive its highest level of funding. His research and commentary have been featured in press outlets including the Washington Post, NPR, The Atlantic, Buzzfeed, Vox, USA Today, the BBC, PBS NewsHour, CBS News, NBC News, and many others. Unlike many computational social scientists, he centers questions of identity and power in his work, paying particular attention to race, gender, and ideology.

The Computational Social Science Lab Speaker Series will be co-hosted by the NetSCI Lab and will feature Presidential Professor Deen Freelon, University of Pennsylvania, who will offer insights into using AI tools to analyze media discourse about racism.

Deen Freelon
Analyzing frames around “racism” in the news media: An AI-assisted approach
Mediated discourse around racism has been a topic of occasional interest in political communication research over the past few decades, but traditional approaches have left vital questions unanswered. Much of this work is qualitative, offering penetrating theoretical insights but little in the way of generalizable results. The quantitative side of the literature is dominated by content analyses—both manual and automated--some of which merge racism with other offensive communications under the umbrella concept of “hate speech.” The current approach focuses specifically on the discursive contexts in which the term “raci(al/sm/st)” (including euphemisms like racial/ly ”charged,” “tinged,” “motivated,” “connotations,” etc.) appears in historical and contemporary mainstream, left-wing, right-wing, and Black news media. In keeping with definitions of framing that focus on problem definition and assignment of blame, this presentation will demonstrate the use of AI (specifically LLMs) to identify claimants, objects, perpetrators, and targets of racism and related concepts in mediated discourse. This project, which is in its initial planning phase, has two main goals: first, to identify the most important predictors of certain groups (e.g., racial, gender, age) appearing as perpetrators and targets of racism; and second, to produce a comprehensive, quantitative impression of how these trends have changed over time. 

Deen Freelon is a Presidential Professor at the Annenberg School for Communication. A widely recognized expert on digital politics and computational social science, he has authored or coauthored over 60 book chapters, funded reports, and articles in journals such as Nature, Science, and the Proceedings of the National Academy of Sciences. He was one of the first communication researchers to apply computational methods to social media data and has developed eight open-source research software packages. The first of these, ReCal, is a free online intercoder reliability service that has been running continuously since 2008 (when he was a Ph.D. student) and used by tens of thousands of researchers worldwide. He has been awarded over $6 million in research funding from grantmakers including the Knight Foundation, the Hewlett Foundation, the Spencer Foundation, and the US Institute for Peace. He was a founding member and remains Senior Researcher at the Center for Information, Technology, and Public Life at the University of North Carolina at Chapel Hill, one of five academic research centers in the Knight Research Network (est. 2019) to receive its highest level of funding. His research and commentary have been featured in press outlets including the Washington Post, NPR, The Atlantic, Buzzfeed, Vox, USA Today, the BBC, PBS NewsHour, CBS News, NBC News, and many others. Unlike many computational social scientists, he centers questions of identity and power in his work, paying particular attention to race, gender, and ideology.

Room 323