Analysis and detection of influence operations and deception in social media

Influence operations of various kinds are often part of modern political and military conflicts. The use of deception and information operations in social media is also heavily increasing, due to the development and widespread use of social media platforms such as Twitter and Facebook. There are many examples of how state actors as well as non-state actors are using social media to influence the national or international opinion. Moreover, terror organizations such as Islamic State have been successful in using social media to spread propaganda and to recruit new members. In this report we present technical solutions for detection of influence operations on the microblog Twitter. The suggested prototype tools make it possible to collect data from open sources and combine data-driven technologies and methods (such as natural language processing and algorithms for detection of automated behavior and hijacked user accounts) with interactive information visualization, enabling a workflow where human and machine can work together in order to identify potential deceptions in the large amounts of data produced every day in social media.