Brandeis Magazine

With support from a $731,000 grant from the U.S. Department of Justice, politics professor Jytte Klausen and a team of undergraduate and graduate researchers are collaborating with Anura Jayasumana, an electrical and computer engineering professor at Colorado State University, to develop a data-driven computational tool that can detect when potential terrorists are on the path to radicalization.

Jayasumana is building an algorithm using information from Klausen’s Western Jihadism Project, a multimedia data archive that is monitoring the growth of jihadism in Europe, North America and Australia since 1993.

More than 80 Brandeis students have contributed to the project since its launch in 2009. The database now holds thousands of records — information about terrorist offenders; jihadist recruitment efforts; social media networks; local and international organizations linked to Western jihadist extremists; and 797 plots that failed or were foiled, or were successfully executed. All the information is painstakingly culled from public sources such as court documents and news reports in many languages, and manually coded into Klausen’s web-based database.

“Our goal is to use real-world data collected from the biographies of known terrorist offenders to develop a dynamic tracking mechanism that can measure increased radicalization,” says Klausen. “My research shows there are predictable sequences in how people move from being interested in an extremist ideology to being capable of taking violent action.” She believes the project could change how law-enforcement personnel and NGOs working to prevent violent extremism assess risks associated with terror threats as well as how social scientists conduct research on extremist groups.

“This is a totally exciting project that puts together new and inventive work in both computer science and the study of political movements,” Klausen says. “The idea of using a dynamic radicalization model has been tested by analysts in multiple countries, and it works.”