Abstract
Hate speech is one of the major concerns of Europe. Different studies, mainly in English language, have been carried out to analyze hate speech, many of them from a theoretical perspective. Here, it is presented an observational study about hate speech poured on Twitter in Spanish regarding to five social important events: Women´s Day, International LGTBQ+ Pride Day, Spain National Day, national elections, and regional elections. Three different experiments were carried out; two used deep learning algorithms to automatically classify tweets, meanwhile, the latest tweets were classified by a human. Results showed that these events significantly triggered hate speech, yet results differed between experiments, and also depending on the nature of the events. A better understanding of the mechanisms of hate speech propagation can help improve policies in Spain or in countries with similar characteristics, and thus help law enforcement and other institutions to address the scourge of hate crimes.
TopIntroduction
Hate crimes in the European Union are a growing concern as they violate fundamental rights protected by the Universal Declaration of Human Rights (Anderson & Meyer, 2016). Social media have facilitated the spread of offensive and hate speech, with users hiding behind the right to freedom of expression (Whitney, 2018; Arcila et al., 2022). Most research has focused on message text analysis in English, with fewer studies in Spanish or other languages (Aluru et al., 2020; Basile et al., 2019; Battistelli et al., 2020; del Valle-Cano et al., 2023; Florio et al., 2017; Pereira-Kohatsu et al., 2019; Plaza-del Arco et al., 2021; Poletto et al., 2021; Sreelakshmi et al., 2020).
In Spain, over 22% of 2021's internet hate crimes occurred on Social media, with rising percentages (López-Gutiérrez et al., 2021). Hate speech on these platforms has been identified as a key driver of hateful attitudes and hate crimes (Amores et al., 2021). Triggered by social events, research has shown a connection between online hate speech and offline hate crime (Williams and Burnap, 2016; Olteanu et al., 2018; Relia et al., 2019; Müller & Schwarz, 2020, 2021; Scharwächter & Müller, 2020; Williams et al., 2020; Arcila et al., 2022). This blurs the boundary between online and offline spaces, making hate speech an extension of the physical environment (Awan & Zempi, 2016). Negative impacts of offensive comments on victims' mental health include increased stress, anxiety, and depression (Baumgartner et al., 2018).
Online platforms have implemented measures to control hate speech, but regulation procedures are not standardized, leading to varied criteria across Social media (Berger & Morgan, 2016). These regulations often stem from platform policies rather than legislation, impacting freedom of expression, privacy, and community building (Gillespie, 2018). Consequently, systems on Social media can reflect and perpetuate social inequalities, especially regarding gender and race, disproportionately affecting vulnerable groups and amplified by platform characteristics (Seeta, 2019).
Key Terms in this Chapter
Twitter: This is a free social network that allows users to stay connected through short text messages of up to 280 characters.
Intolerant Speech: Without being defined as a criminal offence by the penal code, it affects certain groups in a discriminatory manner.
Hate Speech: Some authors refer to hate speech from an offensive and discriminatory point of view, but if the Spanish penal code is taken into account, the definition would be related to the direct or indirect promotion of hatred, although that legislation includes more casuistry.
Natural Language Processing: They are algorithms based on different technologies, such as neural networks, which allow the analytical processing of texts.
Social Important Events: Throughout this chapter, these types of events have been named, taking into consideration those that significantly affect the whole society, whether or not all the individuals belong to a certain group, because their public diffusion is great.
Predictive Policing: Predictive policing allows the analysis of past data to predict behaviors or events that may occur in the future with an associated probability.
Spanish National Office Against Hate Crimes (ONDOD): The main goal of the National Office is to implement measures at the national level to enable law enforcement agencies to respond more effectively to victims of hate crimes and to combat hate crimes more effectively.