MENU

New Computer Algorithm Sifts Social Media Posts To Predict Next ISIS Terror Attack

Close

The ability to identify the signs that precede acts of terrorism that could claim lives is crucial to preventing these incidents from happening again.

Scouring social media posts appears to be a good start for government agencies and terrorism experts to find clues that promote the cause of militant groups or online activities that precede terror acts.

Before opening fire at an Orlando gay nightclub on June 12, for instance, Omar Mateen posted Islamic State-related threats on Facebook.

Now, a physicist and his team may just help improve the potentials of using social media in predicting and possibly preventing the next terror attack.

Neil Johnson and colleagues from the University of Miami have created a mathematical model to figure out how the ISIS terrorist group manages to grow and organize online, which may help thwart future attacks.

The researchers focused on activities on the social platform VKontake, or VK. Johnson explained that they chose the Russia-based social media site because pro-ISIS pages are immediately shut down on Facebook.

The site, which has about 350 million users worldwide, allows use of multiple languages, and according to the researchers, is used by ISIS to spread propaganda among Russian-speaking population.

The researchers looked for pro-ISIS posts daily from mid-2014 up to August 2015, sifting through posts for mentions of activities linked to the militant group in multiple languages. They eventually come up with an equation that shed light on the online activities of Islamic State sympathizers.

In their study published in the journal Science on June 17, the researchers said that the groups with pro-ISIS sentiments have predictable behaviors prior to the attacks.

Analyses revealed that pro-ISIS groups survive like mushrooms. Although groups are occasionally shut down by hackers or online moderators, followers of these groups sprout in new groups or move to another pack.

A sudden increase in the number of ISIS-supporting aggregates were found to likely precede an act of violence.

"The ecology features self-organized aggregates (ad hoc groups formed via linkage to a Facebook page or analog) that proliferate preceding the onset of recent real-world campaigns and adopt novel adaptive mechanisms to enhance their survival," the researchers wrote in their study.

Johnson and colleagues likewise found a potentially efficient way to fight pro-ISIS activities. The computer algorithm revealed that eliminating small groups can help to disable the distribution of the ISIS propaganda.

"One of the predictions is that development of large, potentially potent pro-ISIS aggregates can be thwarted by targeting smaller ones," the researchers added.

See Now: 30 Gadgets And Tech Gifts For Father's Day 2018 That Dad Will Think Are Rad

ⓒ 2018 TECHTIMES.com All rights reserved. Do not reproduce without permission.
Real Time Analytics