Modular Analytics and Statistical Quantification of Adversarial Engagements (MASQUERADE)

Within the past decade, botnets have evolved from simplistic information distributors, which were more of a nuisance than a threat, to complex, weaponized tools that take advantage of cultural values to propagate highly targeted and charged propaganda. Although the ability to detect and map these information campaigns exist, the required capabilities are fragmented across multiple domains of expertise (e.g., knowledge of the campaign, graph and network analytics, and natural language processing). To address the need for analyst-centered tools that identify emerging information campaigns, and allow analysts to keep up with botnets as they evolve over time, we propose to design and demonstrate a Modular Analytics and Statistical Quantification of Adversarial Engagements (MASQUERADE) system. MASQUERADE uses a probabilistic model to maps the complex behaviors expressed by botnets in social networks to the high-level maneuvers carried out by the actors leading for information campaign. MASQUERADE provides analysts the capability to detect information campaigns as they begin, understand the characteristics of a campaign underway, and predict the next steps the campaign will take.