Skip to Content

Data Mining Reveals the Rise of ISIS Propaganda on Twitter

Twitter has closed 25,000 accounts that supported the terrorist organisation ISIS. An analysis of these tweets shows how ISIS emerged with a message of extreme violence.

In the chaos that followed the American withdrawal from Iraq and the popular uprising in Syria, a number of groups began promoting various religious, political, and military agendas. Of all these, one has grown to dominate the region through fanatical religious ideology, sectarianism, and extreme violence.

The so-called Islamic State of Iraq and Levant, or ISIS, as it has become known, is a militant group that follows a fundamentalist doctrine of Sunni Islam. In 2014, it proclaimed itself a caliphate while claiming religious, political and military authority over all Muslims. Through its military activities, ISIS gained control of huge swatches of land in Iraq and Syria while committing numerous war crimes and human rights abuses.

One factor behind this rapid rise to power was ISIS’s use of social media, and Twitter in particular, to spread its ideas. And that raises some interesting questions: what have ISIS members and sympathizers been talking about on Twitter?  And why did that message prove so infectious?

Today we get an answer thanks to the work of Adam Badawy and Emilio Ferrara at the University of Southern California in Los Angeles. These guys have analyzed some two million messages posted on Twitter by 25,000 members of ISIS. They say their analysis reveals important insights into the way the radical militant groups use social media and why its message spread so rapidly.

Identifying ISIS sympathizers on Twitter is not an easy task, not least because much of the discourse is in Arabic. So in 2015, an anonymous collection of activists began the task of manually identifying ISIS-related accounts on Twitter.

The called themselves the Lucky Troll Club and reported all the accounts they identified to Twitter. Twitter than checked each of these accounts itself and suspended 25,000 of them for supporting terrorist-related activity.

Badawy and Ferrara created their data set by collecting all the tweets sent from these 25,000 accounts between January 2015 and June 2015. That’s some 1.9 million messages in total.

Analyzing such a data set is not easy. Badawy and Ferrara began by removing all the tweets that were not in Arabic, about 8 percent of them. They then reduced the Arabic words in the data set to their stem or basic meaning and compiled a list of the 100 most popular stems. Of these, 66 stems had no clear relevance to the meaning of tweets, such as to go out: خرج ; land, ارض ; and to do, ه.

Removing these from the list left them with 34 stems related to acts of violence, theological pronouncements, and sectarianism. Violence related stems include to kill, ق ل ; and bomb, ف قص. Theological stems include Judgement Day, حسب and Caliph خلف. While sectarian stems include كفر and رتد which can be used as derogatory terms for Shias, Yazidis, Christians, and even Muslims who do not adhere to ISIS’s vision of Islam.

Finally, Badawy and Ferrara used a machine-learning algorithm to classify the tweets in the database according to the stems they contained.

The results make for interesting reading. “Theological and violence related issues compose a little over 30 percent of all the tweets,” say Badawy and Ferrara.

That’s a curious finding because it raises the question of why ISIS places such an emphasis on violence? Badawy and Ferrara think they know the answer. Here is part of their discussion:

“Violence plays a major part in ISIS’s brand and its appeal among ISIS followers. We suggest that ISIS transformed the goal of many Islamist groups. Since the foundation of al-Qaeda, Osama Bin Laden advocated focusing on the ‘head of the snake,’ as he called the USA. He is arguing that once one destroys the source of ‘evil’ that dominates Muslim countries, its puppets (Arab and Muslim leaders) would consequently lose power, and Islamic lands would be freed. This message insisted on gradualism and portrayed the battle between Jihadists and the US and its supporters in the context of David vs. Goliath. On the other hand, ISIS, from the start, refused to portray itself as an underdog, focusing on its victories, atrocious violent acts against minorities (and particularly, Shias, whom they used as a target in order to gain support among certain sectors of the Sunni community), and its call for an Islamic state.”

Badawy and Ferrara go on to say this was a deliberate attempt to distinguish ISIS from other groups. “This was an intentional posture, which is reflected profoundly in its message to: 1) excite and attract many young men and women to this ‘exciting’ and ‘victorious’ journey that brings pride to its participants, and 2) to escape the spiritual hegemony of al-Qaeda-central and Ayman al-Zawahiri’s attempts to control the jihadi scene in Iraq and Syria.”

The researchers go on to plot the tweets over time and show how violent and sectarian messages peak in the aftermath of violent actions and sometimes precede them.

That’s interesting work that gives a persuasive answer to the question of why ISIS, and not another group, emerged as the most powerful organization in the region. “A major reason behind that is its spectacular ability to spread its violent and nihilistic messages further and better than any of if its rivals, including established groups like, al-Qaeda,” conclude Badawy and Ferrara.

There is much to be learnt from analyses like this. But how it can be used to prevent these kinds of organizations from emerging in the future is far from clear. If anything, it provides a template for their success. And therein lies an important challenge for the future.

Ref: arxiv.org/abs/1702.02263: The Rise of Jihadist Propaganda on Social Networks

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.