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 --

LEVEL 1. Knowing Online Hate Speech (KNowHATE)

BRIDGE Main Challenge: A.5. Discrimination, racism and sociocultural polarisation, including hate speech


What problem does it address?

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.

What does it propose?

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.

Why is it an inspiring practice?
  •  Combines social sciences, linguistics and artificial intelligence.
  •  Focuses specifically on online hate speech in Portuguese.
  •  Addresses both direct and subtle/indirect forms of hate speech.
  •  Involves communities directly affected by discrimination.
  •  Connects scientific evidence with lived experiences.
  •  Develops AI-based tools for automatic hate speech detection.
  •  Promotes counter-speech rather than focusing only on content removal.
  •  Supports critical digital and media literacy.
  •  Produces accessible resources for educators and civil society.
  •  Translates research findings into practical tools and public awareness actions.
Resources you'll find
  •  KNOwHATE AI Prototype for Online Hate Speech Detection.
  •  Knowledge is Our Weapon – Booklet.
  •  Cortar o Mal Pela Raiz practical leaflet.
  •  Cortar o Mal Pela Raiz awareness campaign.
  •  Cortar o Mal Pela Raiz Podcast Series.
  •  Online awareness videos.
  •  Glossary on Online Hate Speech.
  •  Scientific publications and research results.
Featured results
1
KNOwHATE AI Prototype for Online Hate Speech Detection.
1
Podcast Series: “Cortar o Mal Pela Raiz”.
1
Booklet: “Knowledge is Our Weapon”.
1
Awareness campaign.
Multiple
Online awareness videos, scientific publications and research results.
What can visitors learn?
 How to identify direct and indirect online hate speech.
 How hate speech affects different social groups.
 How stereotypes, irony and humour can convey subtle forms of hate.
 How AI can support automatic hate speech detection.
 How to distinguish hate speech from freedom of expression.
 How counter-speech can challenge discriminatory narratives.
 How language and context influence hate speech detection.
 How social media contributes to the spread of discriminatory content.
 How to respond critically to online hate speech.
 How research can support digital citizenship and media literacy.

LEVEL 2. EXPLORE BEST PRACTICE

Introduction

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.

  • Coordinating organisation: ISCTE – Instituto Universitário de Lisboa
  • Participating countries: Portugal (PT)
  • Duration: February 2022- july 2024
  • BRIDGE Main Challenge: A.5. Discrimination, racism and sociocultural polarisation, including hate speech
Sustainable Development Goals (SDGs)
  • SDG 4: Quality Education
  • SDG 5: Gender Equality
  • SDG 10: Reduced Inequalities
  • SDG 16: Peace, Justice and Strong Institutions
Secondary Challenges
  • A.1. Structural and intersectional inequalities that shape inclusion and social cohesion
  • A.7. Digital ecosystem, disinformation and the need for media literacy and critical thinking

The Challenge

What social issue does it address?

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.

Why is it important?

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.

How does it contribute to the BRIDGE Challenge?

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.

How does it work?

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.

What might inspire you?
Ideas for adaptation
  1. Analyses real examples to recognise direct and subtle online hate speech.
  2. Explores how stereotypes and humour can conceal discriminatory messages.
  3. Discusses the boundaries between hate speech and freedom of expression.
  4. Creates counter-narratives to challenge discriminatory online content.
  5. Uses podcasts to stimulate classroom discussion about online hate.
  6. Explores how AI can help detect hate speech on social media.
  7. Compares hate speech and counter-speech across digital platforms.
  8. Develops critical media literacy using real social media examples.
  9. Discusses how online hate affects different communities unequally.
  10. Turns research evidence into digital citizenship activities.
Resources created

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

https://knowhate.eu/project-results/

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

https://knowhate.eu/project-results/

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

https://knowhate.eu/project-results/

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

https://knowhate.eu/project-results/

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

https://knowhate.eu/project-results/

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

https://knowhate.eu/project-results/