Online Library

Islamophobic hate speech on social media is a growing concern in contemporary Western politics and society. It can inflict considerable harm on any victims who are targeted, create a sense of fear and exclusion amongst their communities, toxify public discourse and motivate other forms of extremist and hateful behavior. Accordingly, there is a pressing need for automated tools to detect and classify Islamophobic hate speech robustly and at scale, thereby enabling quantitative analyses of large textual datasets, such as those collected from social media. Previous research has mostly approached the automated detection of hate speech as a binary task. However, the varied nature of Islamophobia means that this is often inappropriate for both theoretically informed social science and effective monitoring of social media platforms. Drawing on in-depth conceptual work we build an automated software tool which distinguishes between non-Islamophobic, weak Islamophobic and strong Islamophobic content. Accuracy is 77.6% and balanced accuracy is 83%. Our tool enables future quantitative research into the drivers, spread, prevalence and effects of Islamophobic hate speech on social media.

This report aims to contribute to our understanding of online hate in the context of the requirements of the revised Audiovisual Media Services Directive (AVMSD) for Video Sharing Platforms (VSPs) to protect the general public from incitement to hatred or violence. However, online hate is complex and it can only be fully understood by considering issues beyond the very specific focus of these regulations. Hence, we draw on recent social and computational research to consider a range of points outside VSP regulations, such as the impact, nature and dynamics of online hate. For similar reasons,
we have considered expressions of hate across a range of online spaces, including VSPs as well as other online platforms. In particular, we have closely examined how online hate is currently addressed by industry, identifying key and emerging issues in content moderation practices. Our analyses will be relevant to a range of experts and stakeholders working to address online hate, including researchers, platforms, regulators and civil society organisations.

Online abusive content detection is an inherently difficult task. It has received considerable attention from academia, particularly within the computational linguistics community, and performance appears to have improved as the field has matured. However, considerable challenges and unaddressed frontiers remain, spanning technical, social and ethical dimensions. These issues constrain the performance, efficiency and generalizability of abusive content detection systems. In this article we delineate and clarify the main challenges and frontiers in the field, critically evaluate their implications and discuss potential solutions. We also highlight ways in which social scientific insights can advance research. We discuss the lack of support given to researchers working with abusive content and provide guidelines for ethical research.

This study aims only at analysing the dynamics of jihadist radicalisation in Italy. It is not meant to be a study of Islam or of the Italian Muslim community. Rather, it describes a phenomenon that, in Italy as in any other Western country, affects a statistically insignificant percentage of the Muslim population. Throughout Europe, jihadism is a fringe phenomenon, much debated but affecting only a few isolated individuals within largely peaceful Muslim communities.

ƒWhile not as large as in many other Western
countries, ISIS-related mobilization in the United States
has been unprecedented.

As of the fall of 2015, U.S.
authorities speak of some 250 Americans who have
traveled or attempted to travel to Syria/Iraq to join the
Islamic State in Iraq and Syria (ISIS) and 900 active investigations
against ISIS sympathizers in all 50 states.

ƒ Seventy-one individuals have been charged with
ISIS-related activities since March 2014. Fifty-six have
been arrested in 2015 alone, a record number of
terrorism-related arrests for any year since 9/11. Of
those charged:
‚ The average age is 26.
‚ 86% are male.
‚ Their activities were located in 21 states.
‚ 51% traveled or attempted to travel abroad.
‚ 27% were involved in plots to carry out attacks on
U.S. soil.
‚ 55% were arrested in an operation involving an
informant and/or an undercover agent.

ƒ A small number of Americans have been killed in
ISIS-related activities: three inside the U.S., at least a
dozen abroad.

ƒ The profiles of individuals involved in ISIS-related activities
in the U.S. differ widely in race, age, social class,
education, and family background. Their motivations
are equally diverse and defy easy analysis.
ƒ Social media plays a crucial role in the radicalization
and, at times, mobilization of U.S.-based ISIS sympathizers.
The Program on Extremism has identified some
300 American and/or U.S.-based ISIS sympathizers active
on social media, spreading propaganda, and interacting
with like-minded individuals. Some members
of this online echo chamber eventually make the leap
from keyboard warriors to actual militancy.

ƒ American ISIS sympathizers are particularly active on
Twitter, where they spasmodically create accounts that
often get suspended in a never-ending cat-and-mouse
game. Some accounts (the “nodes”) are the generators
of primary content, some (the “amplifiers”) just retweet
material, others (the “shout-outs”) promote newly created
accounts of suspended users.

ƒ ISIS-related radicalization is by no means limited to
social media. While instances of purely web-driven,
individual radicalization are numerous, in several cases
U.S.-based individuals initially cultivated and later
strengthened their interest in ISIS’s narrative through
face-to-face relationships. In most cases online and
offline dynamics complement one another.

ƒ The spectrum of U.S.-based sympathizers’ actual involvement
with ISIS varies significantly, ranging from
those who are merely inspired by its message to those
few who reached mid-level leadership positions within
the group.

This chapter examines Europol’s Internet Referral Unit (EU IRU) and its role in the privatized governance of online media within the broader context of transnational law enforcement and European counter-terrorism policy. Established in 2015, the EU IRU monitors online content and recommends takedowns to hosting platforms based on breaches of terms of service, not legal violations. By analyzing the discursive practices of the EU IRU, this chapter seeks to understand the (in)security dynamics of online radicalization and assess associated human rights concerns. The core question addressed is how Europol frames radicalization and Internet content. The chapter situates the EU IRU within the wider discourse on counter-radicalization, arguing for a deeper understanding of radicalization as perceived by professionals in the field. A detailed policy analysis highlights the unit’s referral and takedown activities, scrutinizing Europol’s framing of technology as a security issue and radicalization as a problem exacerbated by online communication. The study reveals significant issues such as over-blocking, lack of due process, and absence of recourse for affected users, emphasizing the ethical and human rights implications of private companies acting as arbiters of speech. The conclusion critiques Europol’s approach, noting its inconsistency and inefficacy, and underscores the necessity of transparent and judicially overseen content regulation. This chapter calls for a re-evaluation of current counter-radicalization strategies, advocating for more holistic approaches that balance security needs with fundamental rights and democratic accountability.