Predicting and Preventing Extremist Violence with the Power of Quantitative and Visual Analytics

This event is part of our Social and Decision Analytics Seminar Series. To view a six-minute teaser video for Dr. Towers research, click here. 

Speaker: Dr. Sherry Towers, Owner of Towers Consulting, LLC, and Affiliate Scholar at the Institute for Advanced Sustainability Studies, Potsdam, Germany

Bio: Dr. Sherry Towers is a data scientist with a diverse background in visual analytics, data mining, social media analytics, machine learning, high-performance computing, and mathematical and computational dynamical modeling. As both an academic and a consultant, Dr. Towers has worked on a broad array of research topics in public health and the social sciences, including crime and violence risk analyses, the spread of political and partisan sentiments in society, disease modeling, and other topics in applied modeling in the social and life sciences, with over 360 publications on a wide variety of subjects. Her unique trans-disciplinary skill set enables her to examine a wide range of research questions, often of broad interest and importance to policymakers and the general public.

Abstract: 2020 was a climacteric year in the United States, with widespread protests related to the pandemic, police violence, racial issues, and the presidential election. There was an explosion of violence associated with these protests that was often partisan in nature, and 2020 saw a dramatic rise in the participation of armed extremist groups in protests nationwide. Over the past two years, Dr. Sherry Towers has collaborated with partners including Cure Violence Global, the Carter Center and the DHS to better understand the temporal and geospatial trends in extremist and partisan violence in an effort to better direct on-the-ground community resources towards intervention efforts to prevent violent conflict. As she will discuss in this talk, pinpointing which communities are at highest risk of such violence involves a wide variety of data on past protests and violence, political mobilization, socioeconomic demographics, social media data, and data related to extremist recruitment and organizing activities. While quantitative analytics play a significant role in the process, it is critical to also include the "human intel" component, drawing upon the input of experts in partisan violence and extremism to best pinpoint areas for optimal intervention. To facilitate this, the projects have heavily relied upon interactive visual analytics applications that allow these non-data-scientist experts to meaningfully sift through the extremely broad array of potentially relevant data and additionally input their own expertise towards effective decision making.

Join the seminar here.