Data-driven Toolkit to Combat Cyberbullying among Teens (DTCCT)

Doctoral project by Kanishk Verma

Supported by the DCU Anti-Bullying Centre and ADAPT Centre

What is DTCCT ?

We are trying to understand how bullying happens on social media amongst teenagers and we are creating tools to combat such antisocial online behaviour. This project is aligned with the United Nations (UN) Rights of the Child and the recent report by European Commission Joint Research Centre (JRC) advocating for the inclusion of children and young people in design of AI systems. We want to hear from teenagers and young adults who can help us design effective AI systems for moderating bullying-related content. We want to know things like:

  • Is nasty or hurtful content shared on apps like TikTok, Snapchat, and Instagram?
  • How do people share hurtful videos, pictures, and comments on social media?
  • Why do some people encourage bullies or mean people and others try to stop them?

DTCCT is the first attempt at developing and designing AI-based online content moderation toolkit from the perspective of teenagers and young adults.

How to Participate in this research?

You can participate in any of the following research phases to help create an effective Anti-Cyberbullying AI-driven toolkit.

  • Junior Research Assistants: If you are a transition year student or a teenager interested in learning about AI and help in desigining anti-bullying tools, Click here to know more.
  • Specialised Annotator: If you are interested in paid volunteering for data annotation, you can help researchers understand type of Cyberbullying by marking existing images, videos and comments. Click here to know more
Your participation will drive the creation of a novel and innovative toolkit to tackle Cyberbullying.

Research Ethics

This research project has received Ethics Approval from Dublin City University to carry out this research.



Publications

Computational Science Publications


  1. Beyond Binary: Towards Embracing Complexities in Cyberbullying Detection and Intervention - A Position Paper (Conference Article)

  2. Leveraging machine translation for cross-lingual fine-grained cyberbullying classification amongst pre-adolescents (Journal Article)

  3. Benchmarking Language Models for Cyberbullying Identification and Classification from Social-media Texts (Workshop Article)

  4. Can attention-based transformers explain or interpret cyberbullying detection? (Workshop Article)

Social Science Publications


  1. Effectiveness of Artificial Intelligence–Based Cyberbullying Interventions From Youth Perspective (Journal Article)

  2. Examining the Effectiveness of Artificial Intelligence-based Cyberbullying Moderation on Online Platforms: Transparency Implications (Conference Article)

  3. Desigining Ethical Artificial Intelligence (AI) systems with meaningful youth participation: Implications and Considerations (Conference Article)

Team

This project is part of the doctoral thesis of Kanishk Verma, who is a PhD student at the School of Computing, DCU, under the joint supervision of Dr Brian Davis at Dublin City University (DCU), Dr Tijana Milosevic at University College Dublin (UCD), and Dr Rebecca Umbach at Google. This project has received funding from the Irish Research Council (IRC) and Google, Ireland, under the Online Content Safety grant number EPSPG/2021/161

This project is supported by the DCU Anti Bullying Centre and the ADAPT SFI Centre

Kanishk Verma

PhD Student

Dr Brian Davis

Supervisor

Dr Tijana Milosevic

Supervisor

Dr Rebecca Umbach

Enterprise Mentor

Funding Acknowledgemnt

This project has received funding from the Irish Research Council (IRC) and Google, Ireland, under the Online Content Safety grant number ESPSG/2021/161