Online Library

This paper presents a survey on hate speech detection. Given the steadily growing body of social media content, the amount of online hate speech is also increasing. Due to the massive scale of the web, methods that automatically detect hate speech are required. Our survey describes key areas that have been explored to automatically recognize these types of utterances using natural language processing. We also discuss limits of those approaches.

In Part I, this Note will look at the European Union’s current Internet Forum and their proposed Terrorist Content Regulation. The proposed regulation focuses on countering terrorist content on social media by placing much of the burden on the social media companies themselves. Part II will focus on the Material Support of Foreign Terrorist Groups Statue: 18U.S.C.§ 2339(b).

Internet Communication Technologies (ICTs) have become increasingly popular for social movements of all kinds. Social movement organizations (SMOs) use them to interact and mobilize also across borders. Previous research has highlighted the opportunities for transnational networking and the resulting reduction in the importance of local dependency for SMOs. One movement that combines local direct action with the goal of an international network of activists is Antifa. Single case studies have highlighted that local demonstrations and networks are still vital for antifascist groups. However, studies that examine the influence of spatial dynamics on building an international (digital) network beyond focusing on a single country are still lacking. This study, therefore, examines the impact of proximity on forming digital connections among 355 unique European antifascist groups. Using Exponential Random Graph Modeling (ERGM), we examine the hyperlink network across Europe via the groups’ online presence. The findings reveal that the languages of the groups and the geographical distance between them are significant predictors for the probability of displayed hyperlink connections. These results demonstrate the relevance of proximity theories in the field of digital activism. The study contributes to a deeper understanding of spatial dynamics in shaping online networks, reinforcing the importance of space in contemporary network analysis in social movements studies.

Radicalisation can take place offline as well as online. To what extent the internet plays a role remains contested. Doubtless, there are radical and extreme communities online. This report looked into the ability of artificial intelligence (AI) applications to contribute to countering radicalisation. Mapping the possibilities and limitations of this technology in its various forms, the report aims to support decision‑makers and experts navigate the noise, leading to informed decisions unswayed by the current hype.

VOX-Pol Blog post.

To accentuate which platform characteristics particularly foster radicalization and extremist dynamics, this contribution investigates the affordances of social media as delineated in contemporary literature, conducting a platform comparison encompassing Telegram, WhatsApp, and X (formerly Twitter). Based on a scoping review, 17 affordances identified from studies researching radicalization dynamics on social media were extracted and categorized. The most frequently mentioned affordances—anonymity, visibility, and collectivity—were then further analyzed concerning their contribution to radicalization and the radicalization potential of these specific platforms. The platform-comparative affordance discussion shows that, although, in principle each of the three compared platforms can foster an environment conducive to radicalization depending on user intention and usage context, the specific characteristics of each platform necessitate a nuanced consideration. On one hand, it is imperative to discern affordances differentially along various dimensions when assessing their implications (e.g., internal vs. external visibility). On the other hand, it is beneficial to consider which affordances emerge from the actualization of other affordances. For instance, collectivity can result from the interplay of several affordances, such as interactivity and anonymity, and can be referred to as a meta-affordance. Furthermore, the analysis shows that platform branding and self-presentation not only affect platform architecture and affordances but also shape users’ perceptions of the platform, thereby influencing the actualization of affordances. This was particularly noted in the literature for Telegram and increasingly for X. Specific assertions, nevertheless, are hindered by the conceptual diffusion of the affordance approach and a lack of empirical analyses directly and systematically examining platform affordances in conjunction with radicalization dynamics.