Online Library

We present a machine learning framework that leverages a mixture of metadata, network, and temporal features to detect extremist users, and predict content adopters and interaction reciprocity in social media. We exploit a unique dataset containing millions of tweets generated by more than 25 thousand users who have been manually identified, reported, and suspended by Twitter due to their involvement with extremist campaigns. We also leverage millions of tweets generated by a random sample of 25 thousand regular users who were exposed to, or consumed, extremist content. We carry out three forecasting tasks, (i) to detect extremist users, (ii) to estimate whether regular users will adopt extremist content, and finally (iii) to predict whether users will reciprocate contacts initiated by extremists. All forecasting tasks are set up in two scenarios: a post hoc (time independent) prediction task on aggregated data, and a simulated real-time prediction task. The performance of our framework is extremely promising, yielding in the different forecasting scenarios up to 93 % AUC for extremist user detection, up to 80 % AUC for content adoption prediction, and finally up to 72 % AUC for interaction reciprocity forecasting. We conclude by providing a thorough feature analysis that helps determine which are the emerging signals that provide predictive power in different scenarios.

There can be few greater fears for a parent than their child being contacted by a stranger, indoctrinated with an extreme ideology, and encouraged to join a violent movement, all while accessing the internet from their bedroom. Children’s smartphones and computers
may be portals to the most dangerous places on earth. The use of the internet, and more specifically social media, by violent extremists is certainly nothing new. The technical skills and proficiency displayed by groups such as ISIS have been causing concern for governments, law enforcement, industry, schools, religious leaders, and parents around the world. In reality the radicalization of children is rare and particularly nuanced, and far from a linear process that exclusively occurs online. While the media have reported high-profile cases, such as the teenagers from Chicago who tried to leave the United States and join ISIS or the three girls from the United Kingdom who traveled through Turkey to Syria the frequency and nature of conversion to violent extremism is a lot more complex than often reported. Evidence from search histories, online interactions and social media profiles suggest that contact is being made by those intent on radicalizing others. However, while information can be sought and contact can be initiated, complete conversion to a violent ideology is not happening in isolation online. That is to say that the internet is serving as a facilitator, not a direct means of recruitment. In any case action needs to be taken to safeguard children and young people from these risks.

To prevent and counter the spread of illegal hate speech online, in May 2016, the Commission agreed with Facebook, Microsoft, Twitter and YouTube a “Code of conduct on countering illegal hate speech online”. The implementation of the Code of Conduct is evaluated through a regular monitoring exercise set up in collaboration with a network of organisations located in the different EU countries. Using a commonly agreed methodology, these organisations test how the IT companies are implementing the commitments in the Code.

To prevent and counter the spread of illegal hate speech online, in May 2016, the Commission agreed with Facebook, Microsoft, Twitter and YouTube a “Code of conduct on countering illegal hate speech online”. The implementation of the Code of Conduct is evaluated through a regular monitoring exercise set up in collaboration with a network of organisations located in the different EU countries. Using a commonly agreed methodology, these organisations test how the IT companies are implementing the commitments in the Code.

Feminist journalists have come to expect special resistance, and even threats, from men’s groups as part of their work as journalists. However, the biggest threats might not originate in men’s groups’ activities. A big threat currently comes from Internet trolls’ responses to individuals who engage in hate‐provoked and hate‐provoking attacks on women as women. This is exemplified in the case of Anders Behring Breivik, who blew up government buildings in Oslo in 2011 and murdered youth from the Labour Party at Utøya as part of his explicitly articulated xenophobic and misogynist campaign against the Islamification of Norway. His ideas are still being shared in social media responses to this tragedy across Nordic countries. This paper argues that this demonstrates that the harms to women and to society go well beyond the individual victims of an identifiable incident. Largely because of their role in condemning and rejecting the hateful ideas advanced across social media forums, troll responses to the Breivik tragedy constitute a particular threat to female and especially feminist journalists.

Social media systems allow Internet users a congenial platform to freely express their thoughts and opinions. Although this property represents incredible and unique communication opportunities, it also brings along important challenges. Online hate speech is an archetypal example of such challenges. Despite its magnitude and scale, there is a significant gap in understanding the nature of hate speech on social media. In this paper, we provide the first of a kind systematic large scale measurement study of the main targets of hate speech in online social media. To do that, we gather traces from two social media systems: Whisper and Twitter. We then develop and validate a methodology to identify hate speech on both these systems. Our results identify online hate speech forms and offer a broader understanding of the phenomenon, providing directions for prevention and detection approaches.