To facilitate monitoring and research efforts, the sCAN partners took stock of already existing resources and tools. As many of the tools work with keywords to identify hate speech, the sCAN partners compiled a Hate Ontology of keywords in all project languages, including additional information on the context in which those words are used in the respective national discourses. This research provided important insights into the nature of hate speech in the analysed countries.
To complement these findings, the project produced a study mapping available software solutions to automatically monitor cyber hate. The sCAN partners Licra and jugendschutz.net conducted testing campaigns to evaluate the crawlers identified in the mapping study and explore the possibilities of Artificial Intelligence (AI) in monitoring hate speech online. For this, the sCAN project cooperated with the company Factmata, who developed an algorithm to identify ate speech online. The main objective was to provide an evaluation of the accuracy and relevance of the selected tools in order to integrate them in the SCAN consortium monitoring task.
The results of these testing campaigns have been analysed in order to produce a comprehensive user guide on automated monitoring tools. In this guide, the sCAN consortium aims to explain how to use available tools to improve hate speech monitoring and removal.
Furthermore, the sCAN partners conducted joint research projects to analyse specific trends in online hate speech:
- A transnational understanding of the phenomenon of online antigypsyism is necessary to devise effective strategies to counter it. The main narratives of antigypsyism online mirror the historical stereotypes and narratives that have been used for discrimination and persecution of Romani and other communities perceived as ‘gypsies’ for centuries.
- Even though the social media giants Facebook, Twitter and YouTube are most often mentioned in research studies on online hate speech, other platforms are gaining importance especially among young users. Other platforms, like VK.com or Gab.ai are used as alternative platforms or ‘safe havens’ for hate groups or extremist individuals whose profiles have been suspended on mainstream social media. It is therefore necessary to look beyond the “big three” social networks and take a closer look on these alternative platforms.
- Public figures such as politicians, journalists and online influencers can have a strong influence on their followers on social media and in the online sphere. Their wide reach is the reason social media bear special responsibility when it comes to spreading disinformation or implicit (or sometimes explicit) incitement to hatred. In several European countries, leading politicians and other public figures use their online presence to incite hatred or to encourage hate speech by posting biased and populist comments to their social media profiles. Incitement works like a single match causing an entire forest to catch fire, therefore hotspots of online hate need special attention from social media companies.
- By analysing a number of case studies, it was established that intersectional hate speech is common in all countries involved in the project (and probably beyond). Overall, perceived women*, perceived LGBTIQ+ persons and/or persons affiliated or belonging to an ethnic and/or religious minority – based on a combination of their (legally protected) identity categories were identified as the most frequent target groups. Furthermore, people with visible characteristics as well as those in public positions were shown to be particularly affected by intersectional hate speech.
- The Covid-19 pandemic has disrupted human social, economic and political life conditions in all the EU. As the pandemic expands, online phenomena of conspiracy theories, rumours, fake news and hateful contents connected to this global disease are growing.
During the two years of the sCAN project, the partners also participated in four monitoring exercises with the European Commission and the International Network Against Cyber Hate (INACH). The result of these monitoring exercises can be found in our monitoring reports:
- The first monitoring report analyses the sCAN results of the monitoring exercises in November – December 2018 and May – June 2019.
- The second monitoring report details the findings of our monitoring exercises in November – December 2019 and January – February 2020.
Since monitoring and research alone are not sufficient to tackle online hate speech, the sCAN project developed educational training courses both online and offline. One of the sCAN project activities allowed for the Facing Facts! Online Course on online hate speech to be translated into German and French and for adjusting it to the respective national contexts. This course was developed for anyone interested in combatting hate speech online according to their possibilities and capacities. This course offers new insights and practical approaches to effectively countering hate speech online for a broad range of people such as individual activists, members of communities, representatives of CSOs or authorities.
Building on the general online course on hate speech, the sCAN project developed an online course on moderation of hate speech online. This course is available in English and French on the platform Facing Facts Online!. It is directed at activists, community leaders, blogger, vlogger and practitioners interested in encouraging peaceful online exchanges, but any actor of interested in the subject can follow the course.
An offline training on advanced monitoring and countering online hatred was developed and implemented throughout the project duration. Participants had the possibility to become experts in the field of monitoring hate speech and counteraction, documenting the phenomenon, tackling underreporting, comparing results when it comes to data acquired throughout different monitoring exercises and phases, as well as applying effective human rights reporting.
Through our activities we gathered valuable experiences and collected ideas for improvement. The project provided policy recommendations for the institutions of the European Union, national authorities and public institutions, politicians and public figures, social media companies, media and journalists as well as CSOs and individual internet users on how to better combat all forms of hate speech online.
As the internet is not limited by national borders, more transnational cooperation is needed among all stakeholder groups to find a joint approach to this problem.