Online Library

VOX-Pol Blog Post.

Terror attacks are followed by public shock and disorientation. Previous research has found that people use social media to collectively negotiate responses, interpretations, and sense-making in the aftermath of terror attacks. However, the role of ideologically motivated discussions and their relevance to the overall discourse have not been studied. This paper ad-dresses this gap and focuses specifically on the far-right discourse, comparing it to the general public Twitter discourse following the terror attack in Hanau in 2020. A multi-method ap-proach combines network analysis and structural topic modelling to analyse 237,000 tweets. We find responsibility attribution to be one of the central themes: The general discourse pri-marily voiced sympathy with the victims and attributed responsibility for the attack to far-right terror or activism. In contrast, the far right – in an attempt to reshape the general narra-tive – raised a plethora of arguments to shift the attribution of responsibility from far-right activism towards the (political) elite and the personal circumstances of the shooter. In terms of information sharing and seeking, we demonstrate that new information was contextualised differentially depending on the ideological stance. The results are situated in the scientific dis-course concerning differences in social media communication ensuing terrorist attacks.

Social media messages often provide insights into offline behaviors. Although hate speech proliferates rapidly across social media platforms, it is rarely recognized as a cybercrime, even when it may be linked to offline hate crimes that typically involve physical violence. This paper aims to anticipate violent acts by analyzing online hate speech (hatred, toxicity, and sentiment) and comparing it to offline hate crime. The dataset for this preregistered study included social media posts from X (previously called Twitter) and Facebook and internal police records of hate crimes reported in Spain between 2016 and 2018. After conducting preliminary data analysis to check the moderate temporal correlation, we used time series analysis to develop computational models (VAR, GLMNet, and XGBTree) to predict four time periods of these rare events on a daily and weekly basis. Forty-eight models were run to forecast two types of offline hate crimes, those against migrants and those against the LGBT community. The best model for migrant crime achieved an R² of 64%, while that for LGBT crime reached 53%. According to the best ML models, the weekly aggregations outperformed the daily aggregations, the national models outperformed those geolocated in Madrid, and those about migration were more effective than those about LGBT people. Moreover, toxic language outperformed hatred and sentiment analysis, Facebook posts were better predictors than tweets, and in most cases, speech temporally preceded crime. Although we do not make any claims about causation, we conclude that online inflammatory language could be a leading indicator for detecting potential hate crimes acts and that these models can have practical applications for preventing these crimes.

This document reports findings from a Rapid Evidence Assessment conducted on the role of mass and social media during and after terrorist events. It is designed to bring together and synthesize insights and evidence from the available published research literature to inform future policy and practice development. By promoting understanding of how different forms of mediated communication shape what happens in the aftermath of terror events, the work seeks to reflect changes in both the conduct of terrorism and the contemporary information environment. In particular, the spread of social media has had disruptive and transformative impacts upon press and broadcast journalism, and the ways
that terrorist violence is performed.

What explains online radicalization and support for ISIS in the West? Over the past few years, thousands of individuals have radicalized by consuming extremist content online, many of whom eventually traveled overseas to join the Islamic State. This study examines whether anti-Muslim hostility might drive pro-ISIS radicalization in Europe. Using new geo-referenced data on the online behavior of thousands of Islamic State supporters in France, the United Kingdom, Germany, and Belgium, I study whether the intensity of anti-Muslim hostility at the local (neighborhood/municipality) level is linked to pro-ISIS radicalization on Twitter. Results show that local-level measures of anti-Muslim animosity correlate significantly and substantively with indicators of online radicalization, including posting tweets sympathizing with ISIS, describing life in ISIS-controlled territories, discussing foreign fighters, and expressing anti-West sentiment. High-frequency data surrounding events that stir anti-Muslim hostility — terrorist attacks and anti-Muslim protests in Europe — show the same pattern.

Considerable time has been spent examining how groups like AQAP and ISIS used their online magazines to reach and radicalize individuals in Western democratic states. This paper continues this investigation but shifts its analysis to focus on the ‘how-to’ or instructional content of these publications, an understudied part of the literature. One of the stated goals of these magazines was to provide tactical know-how and assist supporters conducting terror plots in their home states. The question: did the tactics outlined in the magazines materialize in actual plots/attacks and how quickly were they put into practice? The paper examines this question by creating an overview of the tactics which appear in these publications and cross referencing them with a dataset of 166 Islamist-inspired homegrown terror plots/attacks in 14 Western democratic states to determine if, and when, they first appeared in relation to their publication date. It concluded that while some of the suggested strategies did appear following their publication, often it occurred after considerable time had elapsed. This suggests the instructional content did not resonate with readers in real time.