Online Library

Countering online hate speech is a critical yet challenging task, but one which can be aided by the use of Natural Language Processing (NLP) techniques. Previous research has primarily focused on the development of NLP methods to automatically and effectively detect online hate speech while disregarding further action needed to calm and discourage individuals from using hate speech in the future. In addition, most existing hate speech datasets treat each post as an isolated instance, ignoring the conversational context. In this paper, we propose a novel task of generative hate speech intervention, where the goal is to automatically generate responses to intervene during online conversations that contain hate speech. As a part of this work, we introduce two fully-labeled large-scale hate speech intervention datasets collected from Gab and Reddit. These datasets provide conversation segments, hate speech labels, as well as intervention responses written by Mechanical Turk Workers. In this paper, we also analyze the datasets to understand the common intervention strategies and explore the performance of common automatic response generation methods on these new datasets to provide a benchmark for future research.

VOX-Pol Blog Post.

Though aiming to remedy genuine harms, government regulation of our online lives also raises legitimate concerns over privacy and freedom of expression. We must address online harms whilst ensuring harms are not also inflicted through unfairly infringing on people’s freedoms. HOPE not hate recognises the importance of this balancing act, and encourages a form of regulation of platforms that places democratic rights front-and-centre. In a world increasingly infused with the web, the significance of this legislation cannot be overstated and it is undoubtedly the case that getting it right will take rigorous reflection. To that end, we encourage debate of the recommendations proposed in this report.

VOX-Pol Blog post.

The Redirect Method​ uses Adwords targeting tools and curated YouTube videos uploaded by people all around the world to confront online radicalization. It focuses on the slice of ISIS’ audience that is most susceptible to its messaging, and redirects them towards curated YouTube videos debunking ISIS recruiting themes. This open methodology was developed from interviews with ISIS defectors, respects users’ privacy and can be deployed to tackle other types of violent recruiting discourses online.