Man Hunt via « Social Big Data »: Use of Social Network Analysis to localize a terrorist

       In line with the massive growth of technology, we are now able to fight more effectively against terrorism by tracking a terrorist network or a potential dangerous individual (isolated node) through a specific social network analysis. Deep learning, Big Data and video surveillance are now the most powerful and effective techniques to follow a possible criminal and to keep track of their moves, to predict future attacks and to unveil their overall network (unrelated, direct, indirect with or without strong ties).

This article will focus on terrorist tracking through social media, introducing a tool: Snaptrends. Snaptrends is a start-up company based in Texas, which offers top-of-the-art software using "social media intelligence". It gathers information with open source intelligence, global tracking and defined geographical position of individuals or influencers through in-depth social network analysis. All the available data are public but the way to treat them is highly innovative. 

As it is a complex tool, it is natural to wonder how it works. Firstly, once the Snaptrends interface is launched, one searches by entering keywords (eg. "ATTACK" or “BOMB”). Then, the application initiates a search by scanning all social feeds (Twitter, Facebook, Google+, Instagram, etc ...) by cross referencing its analysis with another parameter called "geofencing". It actually combines both geolocation and fencing, creating borders, to define a virtual geographic zone. This zone is set by the contents of messages or posts exchanged on social networks: those used directly by the targeted person to make the analysis of related alternative networks. As an example, I ran this App via their official website (http://snaptrends.com) by typing the keyword “Superbowl” and chosing to cross check it using Boston as the location. In this example we can immediately see all the related articles, social feeds, comments and tweets. 

In this example we can immediately see all the related articles, posts, social feeds, comments and tweets. 

Furthermore, when tracking a particular individual, it is very easy to find information about their general profile combined with geolocation. Thus, we have access to all messages sent via this profile on the same day, the day before or the week before. It is also possible to track "back" the activity of the targeted individual by restoring their past messages and to precisely geo-localize all his moves (names of hotels, street address, etc…) and creating a time line. 

The App also permits to analyse a target’s overall network: people who follow him, with whom the target is in direct contact or simply follows. It is also possible to go further by restoring their entire network, using a visualization technique: parabolic graphs (see picture below). 

To conclude, Snaptrends’ main asset lies within its speed and responsiveness and in their highly relevant and advanced geofencing algorithms. The information comes in real time, instantly, which allows up to date searches. In addition, the algorithms analyse not only the networks directly used by the target, but it also takes all related networks into account which makes for a true in depth "Mass Data" analysis. The smart and practical interface allows for the available public data in open sources to be intuitively browsed by anyone and provide according graphs of the target search and their network. A digital tool as developed as Snaptrends combined with other systems should prove to be extremely effective in the fight against terrorism.