RHETORIC

RHETORIC will provide tools for news editors and consumers to detect and counter polarisation on social media and support a civil discourse through semi-automatic moderation of textual and visual comments.

The internet is an important space for democratic debate, but social media have become toxic and polarized, making it difficult for media channels to manage online comments. The RHETORiC project will design interface elements and use machine learning approaches to create semi-automated tools for news editors and media audiences to identify and fight polarization and contribute constructively to online debates.

New forms of expression thwart effective comment moderation

Hateful and offensive comments and reactions to articles have forced many content publishers in Belgium to disable comments from readers. On the other hand, attempts by news channels and authorities to encourage thoughtful dialogue have proven ineffective. With the advent of social media sharing as well as memes and other visual commentary, it has become increasingly difficult to automatically moderate comments to provide safe, open and respectful arenas for online debate.

An outlet for balanced political debate

The RHETORiC project focuses on two main use cases: young people between 16 and 25 years old (VRT NWS case) and the wider audience of Het Niewsblad readers (Mediahuis). For young people,, as research has demonstrated that guidance during this phase promotes tolerance and perspective in political communication. Belgian news group VRT found that young people interested in becoming more politically active lack constructive platforms for useful interaction and balanced debate. RHETORiC seeks to support young people and more generally all Flemish citizens in formulating constructive arguments by developing digital tools based on research and best practices in human-computer interaction design.

Empowering effective moderation and commenting practices

The multidisciplinary RHETORiC consortium consists of experts in media publishing, text and image analytics human-computer interaction and online citizen participation. There are 3 key innovation goals for this project:

  1. Design interface elements that present biases and offer templates for constructive argumentation.
  2. Enable effective semi-automatic comment moderation tools for both text and images.
  3. Create participatory news formats that stimulate online audience conversation.

RHETORiC will develop a demonstrator composed of a moderation dashboard for media channels and commenting tools for social media users, that will be connected with the editorial flow of VRT NWS and Het Nieuwsblad.

Proven value in real-life scenarios of audiene conversation on news topics

The demonstrator created by the RHETORiC project will provide new insights in how to effectively promote helpful civil participation in online debates. User tests will be performed to measure quantitatively and qualitatively how the RHETORiC tools improve the experience, equanimity and balance of debate.

β€œThe RHETORiC project will design interface elements and use machine learning approaches to create semi-automated tools for news editors and media audiences to identify and fight polarization and contribute constructively to online debates.”

RHETORIC

Reducing Hate with Editorial Tools for Online Reactions and Comments.

 

RHETORIC is an imec.icon research project funded by imec and Agentschap Innoveren & Ondernemen.

 

It started on 01.10.2019 and is set to run until 30.09.2021.

Download as pdf

Project information

Industry

  • Textgain
  • Tree
  • VRT
  • Mediahuis
  • Wieni

Research

  • imec – IDLab Data Science Lab – UGent
  • imec - mintlab - KU Leuven

Contact

  • Project lead: Luk Overmeire
  • Research lead: Peter Lambert
  • Proposal Manager: David Geerts
  • Innovation manager: Eric Van der Hulst

Share this on

truetrue

This website uses cookies for analytics purposes only without any commercial intent. Find out more here. Our privacy statement can be found here. Some content (videos, iframes, forms,...) on this website will only appear when you have accepted the cookies.

Accept cookies