Online Library

Societal crises, such as COVID-19, produce societal instability and create a fertile ground for radicalization. Extremists exploit such crises by distributing disinformation to amplify uncertainty and distrust. Based on these developments, this study presents a longitudinal analysis across three different non-violent extremist ideologies in the UK (Islamist, far right and eco-radicals). As part of the study, public social media channels Twitter/X, Facebook, and Telegram) of Britain First, 5 Pillars, and Earth First! were analyzed using a computational language classifier of over 36,000 posts between 2016 and 2021. The increasing prevalence of conspiracy narratives, as well as violent, hateful, and threatening language among the corpus indicates that radicalization dynamics were present and heightened, pre-, during, and post-Pandemic.

Psychological theories of mobilization tend to focus on explaining people’s motivations for action, rather than mobilization (“activation”) processes. To investigate the online behaviors associated with mobilization, we compared the online communications data of 26 people who subsequently mobilized to right-wing extremist action and 48 people who held similar extremist views but did not mobilize (N = 119,473 social media posts). In a three-part analysis, involving content analysis (Part 1), topic modeling (Part 2), and machine learning (Part 3), we showed that communicating ideological or hateful content was not related to mobilization, but rather mobilization was positively related to talking about violent action, operational planning, and logistics. Our findings imply that to explain mobilization to extremist action, rather than the motivations for action, theories of collective action should extend beyond how individuals express grievances and anger, to how they equip themselves with the “know-how” and capability to act.

The purpose of this scoping review is to highlight the machine learning tools used in research to address and prevent violent extremism. To achieve this goal, the following objectives guide this study: (1) describe outcomes that have been studied; (2) summarize the data sources used; and (3) determine whether the reporting of machine learning predictive models aligns with the established reporting guidelines for reporting of prediction models. ProQuest, Compendex, IEEE, JStor and PubMed were searched from June to July 2022. Based on the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) guidelines, databases were searched for articles related to machine learning models applied to the address and prevention of violent extremism. Following standards established by reporting guidelines, findings were extracted from published articles, including general study characteristics, aspects of model development, and reporting of results. Of 53 unique articles identified by the search, 18 were included in the review. Most articles were published between 2016 and 2022 (n = 16, 88.8%). Studies focused on violent extremism worldwide, with the majority of studies not specifically focused on a distinct region (n = 11, 61.1%). The most frequently used machine learning algorithms were support vector machines (n = 9, 50%), followed by random forests (n = 5, 27.7%), natural language processing (n = 4, 22.2%), and deep learning (n = 4, 22.2%). The number of features used varied greatly, ranging from 17 to 7556. Many studies did not report an epistemological or theoretical framework which guided their machine learning approaches or interpretation of findings (n = 8, 44.4%). Many studies did not incorporate the TRIPOD or any other recommended guidelines for the reporting of predictive models. Future research in this field should prioritize evaluating the impact of prediction models on decisions for addressing and preventing violent extremism.

Institutional ethics review procedures aim – in principle – to minimise harm and evaluate risks, providing an important space to consider the safety of participants and researchers. However, literature has questioned the effectiveness of the process, particularly for reviewing ‘risky’ topics in a risk-averse environment. This article reports the findings of interviews with 21 researchers of the far right and manosphere to understand how early career researchers perceive and engage with the process as a component of risk management. It argues that scholars experience IRBs struggling to meet their normative goal of ‘no undue harm’ due to a focus on legality and liability whilst lacking topical and methodological expertise. The lack of expertise produced misperceptions of risk, establishing institutional ethics as an obstacle rather than evaluative aid, creating holes in the ‘safety net’ that institutional ethics can provide. These findings contribute to concerns raised about the effective management of risk by early career researchers and the ethical review of ‘sensitive’ topics.

While certain areas of the Islamic State’s activities (propaganda, recruitment, etc.) are well researched, there have been few studies covering the efforts of the organization to neutralize deplatformization, even though its inclusion in a unified system makes it possible to successfully fight against the organization. The present study investigated the Tactics, Techniques and Procedures (TTPs) applied by the Islamic State to maintain its online presence against the deplatforming efforts of law enforcement agencies and social media stakeholders. A closely related topic, namely the key components of the organization’s successful internet activity, is also examined in this article. During the research, authors mapped and plotted one part of the Islamic State’s internet ecosystem to discover the pattern in this interconnected network. Furthermore, the authors elaborated on a volunteer computing based recommendation for paralysing Islamic State-linked websites.

Most analyses of far-right communication on social media focus on one specific platform, while findings are generalized. In this study, I argue that the far right’s use of social media depends on technology affordances – the linkage between platform design and usage – and, thus, might not always be generalizable. After discussing six affordances – persistence, scalability, replicability, connectivity, searchability, and identifiability – I apply them to studies about the far right on Facebook, Twitter, Instagram, YouTube, and Telegram in a systematic literature review. I conclude that most research focuses on aspects linked to connectivity, scalability, and replicability, while persistence, searchability, and identifiability are barely researched so far. I further discuss shortcomings and possibilities for future research to consider aspect of technology affordances.