Physicist Creates Algorithm To Predict Future ISIS Attacks
A physicist from the University of Miami, straying from his usual field of work, has created an algorithm that can scan social networks to pick up online activity pointing towards a future terrorist attack.
Dr Neil Johnson described how it works in the journal Science Today. The algorithm searched through a year’s worth of posts on the Russian social network Vkontakte, picking up Pro-ISIS conversations in multiple languages. The data then built a statistical model that can identify behavioral patterns among ISIS supporters online.
The study observed the activities of about 100,000 members of 200 different small groups on the Russian site, where the Islamic State has a strong online presence and were sharing their declarations on extremism and ways to survive drone attacks.
The researchers concluded radicalized lone wolves do not stay isolated for long and seek out such online groups eventually. The small groups, because of scathing attacks by the social media themselves, do not tend to last for long and merge with larger groups to expand their influence.
However, the algorithm sure has a lot of shortcomings and the utility of Dr Johnson’s data may be questioned in the line of predicting attacks carried out by a handful of terrorists, for example, the attack in Paris. Retroactive analysis of the data proved the only attack that could have been predicted in the last two years, was the unexpected seige of Kobani in Syria, in 2014. The attack was a military-scale operation by the Islamic State militants and needed social media to gain support and spread propaganda.
Image Source: LA Times
Also Read: Researchers Think Math and Social Media Can Protect Us From ISIS