Pathways to Recruitment

Identifying and comparing social indicators of gang violence and recruitment in a Chicago Twitter data set with ISIS engagement using qualitative analysis and natural language processing.

This project takes an interdisciplinary approach to studying offline social indicators (e.g. precipitating environments and situation events and factors) and online content (e.g. text data, video, images, mood, emotion) emblematic of broader gang behavior and recruitment. This project uses a unique Twitter data set composed of individuals that are embedded in the Twitter network of a well-known deceased Chicago gang member.

We use two methodological approaches: qualitative analysis and natural language processing. The first approach uses in-depth qualitative methods to systematically characterize and associate offline events to social media content and will identify themes and patterns in social media data that can be used to describe offline and online social indictors that describe gang behavior and recruitment. Narrative description is used to compare differences and similarities between gang and ISIS recruitment with the goal of assessing how qualitative differences correlate to quantitative metrics and indicators and how qualitative insight can improve them. Second, we continue work on improving precision of our Natural Language Processing gang violence system for identifying aggression in online posts.

We will broaden this work by adapting it to hold for a broader range of gang members. We will also work on determining its applicability to ISIS social media posts, evaluating it on social media collections. Depending on its performance, we will work on an option to package our aggression indicator and integrate as part of DARPA’s system. We aim to determine the feasibility of automatic correlation of aggression identification with offline violence, transitions between in-group and out-group status, and methods used to recruit members into the in-group. The work is expected to have various potential applications, both for identifying new social indicators of ISIS recruitment on Twitter and curbing gang violence that is initiated on social media.