The project Bridge is co-funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the Servicio Espanol para la Internacionalizacion de la Educacion (SEPIE). Neither the European Union nor the National Agency SEPIE can be held responsible for them. -- Number: 2025-1-ES01-KA220-HED-000359236 --
The BRIDGE Digital Museum brings together good practices, educational proposals and images that inspire reflection and action
BRIDGE Main Challenge: A.5. Discrimination, racism and sociocultural polarisation, including hate speech
The KNOwHATE project addresses the growing spread of online hate speech in Portuguese, particularly against racialised groups, migrants and LGBTI+ communities. Social media facilitates the rapid circulation of direct and subtle forms of discriminatory discourse (indirect hate speech). Although existing, detection tools often fail to capture linguistic, cultural and contextual specificities.
KNOwHATE combines social and language sciences with artificial intelligence to understand, detect and combat online hate speech in Portuguese. It analyses social media content, develops automatic detection models and builds evidence-based, culturally sensitive tools and counter-narratives.
KNOwHATE is an interdisciplinary project addressing online hate speech in Portuguese. Combining social psychology, linguistics, communication, data science and artificial intelligence, it analyses hate speech and counter-speech on social media, develops automatic detection models and creates culturally sensitive resources to support prevention, critical digital literacy and informed responses to discrimination.
Online hate speech reinforces discrimination and structural inequalities, particularly affecting racialised people, migrants and LGBTI+ communities. Digital platforms enable discriminatory messages to spread rapidly, while subtle forms involving stereotypes, irony or humour can be difficult to recognise and detect automatically.
Online hate speech can normalise prejudice, silence marginalised communities and reinforce social inequalities. Developing the capacity to recognise both explicit and subtle forms of hate is essential to protect vulnerable groups, promote critical digital citizenship and strengthen inclusive democratic participation.
KNOwHATE directly addresses discrimination, racism and sociocultural polarisation by generating evidence on online hate speech and developing culturally sensitive detection and prevention tools. Its participatory approach gives visibility to targeted communities and promotes counter-speech, critical literacy and informed responses to discrimination.
The project combines large-scale social media data collection with natural language processing, machine learning and critical discourse analysis. AI detection models are developed and validated alongside qualitative research and input from civil society organisations, ensuring that technological tools reflect the linguistic context and lived experiences of targeted communities.
Knowledge is Our Weapon – KNOwHATE Booklet
Digital publication
An accessible educational booklet presenting the main findings of KNOwHATE and explaining the complex nature of online hate speech in Portuguese. It includes examples and explores direct and subtle (indirect) forms of hate, providing educators, media literacy professionals, policymakers and civil society with knowledge to identify and address discriminatory discourse.
Teaching staff, Students, Researchers, Professionals, Policy Makers, General Public
Cortar O Mal Pela Raiz – Awareness and Educational Resources
Educational resource
An awareness campaign translating KNOwHATE research into accessible public resources explaining what online hate speech is, how it manifests and how it can be challenged. The campaign includes a booklet, practical leaflet, podcasts and audiovisual materials designed to strengthen awareness and critical digital literacy.
Teaching staff, Students, Professionals, General Public
KNOwHATE AI Prototype for Online Hate Speech Detection
Interactive digital resource
An artificial intelligence prototype developed to automatically identify online hate speech in Portuguese. The tool demonstrates how computational models can support the detection of discriminatory discourse while highlighting the importance of linguistic and cultural context.
Teaching staff, Students, Researchers, Professionals
Cortar o Mal pela Raiz – Podcast Series
Educational resource
A podcast series exploring different dimensions of online hate speech, including the boundaries of freedom of expression, different manifestations of hate and links between far-right discourse and online hate. It provides an accessible resource for discussion, awareness and media literacy activities.
Teaching staff, Students, Professionals, General Public
Scientific Publications and Project Results
Digital publication
A collection of scientific publications presenting the project’s research on automatic hate speech detection, social network analysis, counter-speech and linguistic patterns. These resources provide the scientific evidence underpinning KNOwHATE’s technological and educational outputs.
Teaching staff, Researchers, Professionals, Policy Makers
Preliminary Results – Hate Speech on YouTube and X
Digital publication
An accessible presentation of findings from the analysis of tens of thousands of Portuguese-language YouTube comments and Twitter/X posts, highlighting the prevalence, characteristics and differences between direct hate speech, indirect hate speech and counter-speech across platforms.
Teaching staff, Students, Researchers, Professionals, Policy Makers, General Public