This research project produced an integrated computational technology capable of mining, monitoring, and screening for the occurrence of behaviors associated with dangerously escalating extremism, using large heterogeneous databases to identify early warnings of individuals or groups on behavioral trajectories toward extremist violence.
The research succeeded in producing improvements in the analysis methodology of radicalization pathways, along with software to facilitate implementation of the methodology. A library of software routines and computer-aided tools was developed to aid in the compilation of information from documents and for radicalization-indicator detection. This was done by integrating new advances in data science techniques with an evidence-based behavioral model of radicalization identified as leading to violence and terrorism-related crimes. The project focused on three key elements of such behavior: 1) an understanding of extremist radicalization as a process during which the radicalizing individual manifests overt behavioral changes; 2) the development of a trained computational algorithm capable of detecting structured and unstructured data known to signify such behaviors; and 3) a method for analyzing and visualizing the data. Instead of attempting to profile and identify at-risk populations, the current study focused on producing a dynamic, evidence-based assessment model of radicalization trajectories of home-grown militants inspired by the jihadist ideology. Three types of data were collected: 1) demographic data about known jihadist terrorism offenders from the United States and the United Kingdom and a small case-control dataset composed of the so-called “Involuntary Celibate;” 2) information about observed behavioral changes and the timing of these behaviors; and 3) a large collection of text data indicative of cues to such behaviors. 10 figures, 5 tables, and a listing of project-related publications