There is a concerted effort by researchers to understand how the Islamic State of Iraq and Levant (ISIL) is capable of influencing and radicalizing socially vulnerable audiences around the world via digital means. These efforts are demonstrated in a limited body of research that are often times rooted in conventional processes, therefore, having limited direct application to today’s dynamic, open-source digital environment. This environment affords a challenging, yet unique, opportunity to employ open source machine learning techniques guided by social learning and routine activities theory from the criminological field of study. This presentation will discuss a human driven, but machine assisted framework for identifying ISIL methods and victims in order to facilitate an effective counter-narrative for engaging the victims prior to influence happening. The framework utilizes historically based research designs to develop the frameworks, but machine learning to train classification algorithms utilizing data pulled from the Twitter API for modern application. The Scikit-Learn set of tools for Python were used to rapidly prototype tools for data mining and data analysis.