# DoDo Learning: DOmain-DemOgraphic Transfer in Language Models for Detecting Abuse Targeted at Public Figures

arXiv: https://arxiv.org/abs/2307.16811 · HTML: https://mlsanity.com/paper/2307.16811/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 25 of its 105 references that are in the map.
- First posted on arXiv 31 Jul 2023.

## Builds on (most cited first, 25 of 25)

- [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://mlsanity.com/paper/1910.01108/) (2019)
- [Don't Stop Pretraining: Adapt Language Models to Domains and Tasks](https://mlsanity.com/paper/2004.10964/) (2020)
- [DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharing](https://mlsanity.com/paper/2111.09543/) (2021)
- [Automated Hate Speech Detection and the Problem of Offensive Language](https://mlsanity.com/paper/1703.04009/) (2017)
- [Understanding Dataset Difficulty with V-Usable Information](https://mlsanity.com/paper/2110.08420/) (2021)
- [Learning from the Worst: Dynamically Generated Datasets to Improve Online Hate Detection](https://mlsanity.com/paper/2012.15761/) (2020)
- [HateCheck: Functional Tests for Hate Speech Detection Models](https://mlsanity.com/paper/2012.15606/) (2020)
- [Directions in Abusive Language Training Data: Garbage In, Garbage Out](https://mlsanity.com/paper/2004.01670/) (2020)
- [Towards generalisable hate speech detection: a review on obstacles and solutions](https://mlsanity.com/paper/2102.08886/) (2021)
- [Hate Lingo: A Target-based Linguistic Analysis of Hate Speech in Social Media](https://mlsanity.com/paper/1804.04257/) (2018)
- [Hatemoji: A Test Suite and Adversarially-Generated Dataset for Benchmarking and Detecting Emoji-based Hate](https://mlsanity.com/paper/2108.05921/) (2021)
- [SemEval-2023 Task 10: Explainable Detection of Online Sexism](https://mlsanity.com/paper/2303.04222/) (2023)
- [Handling and Presenting Harmful Text in NLP Research](https://mlsanity.com/paper/2204.14256/) (2022)
- [Large-Scale Hate Speech Detection with Cross-Domain Transfer](https://mlsanity.com/paper/2203.01111/) (2022)
- [Hate Speech Detection: A Solved Problem? The Challenging Case of Long Tail on Twitter](https://mlsanity.com/paper/1803.03662/) (2018)
- [Tackling Online Abuse: A Survey of Automated Abuse Detection Methods](https://mlsanity.com/paper/1908.06024/) (2019)
- [Robust Hate Speech Detection in Social Media: A Cross-Dataset Empirical Evaluation](https://mlsanity.com/paper/2307.01680/) (2023)
- [Drinking from a Firehose: Continual Learning with Web-scale Natural Language](https://mlsanity.com/paper/2007.09335/) (2020)
- [Lifelong Learning of Hate Speech Classification on Social Media](https://mlsanity.com/paper/2106.02821/) (2021)
- ["Short is the Road that Leads from Fear to Hate": Fear Speech in Indian WhatsApp Groups](https://mlsanity.com/paper/2102.03870/) (2021)
- [Generalizing Hate Speech Detection Using Multi-Task Learning: A Case Study of Political Public Figures](https://mlsanity.com/paper/2208.10598/) (2022)
- [A large-scale crowdsourced analysis of abuse against women journalists and politicians on Twitter](https://mlsanity.com/paper/1902.03093/) (2019)
- [Does Campaigning on Social Media Make a Difference? Evidence from candidate use of Twitter during the 2015 and 2017 UK Elections](https://mlsanity.com/paper/1710.07087/) (2017)
- [Online Abuse of UK MPs from 2015 to 2019: Working Paper](https://mlsanity.com/paper/1904.11230/) (2019)
- [MP Twitter Abuse in the Age of COVID-19: White Paper](https://mlsanity.com/paper/2006.08363/) (2020)

## Abstract (arXiv)

Public figures receive a disproportionate amount of abuse on social media, impacting their active participation in public life. Automated systems can identify abuse at scale but labelling training data is expensive, complex and potentially harmful. So, it is desirable that systems are efficient and generalisable, handling both shared and specific aspects of online abuse. We explore the dynamics of cross-group text classification in order to understand how well classifiers trained on one domain or demographic can transfer to others, with a view to building more generalisable abuse classifiers. We fine-tune language models to classify tweets targeted at public figures across DOmains (sport and politics) and DemOgraphics (women and men) using our novel DODO dataset, containing 28,000 labelled entries, split equally across four domain-demographic pairs. We find that (i) small amounts of diverse data are hugely beneficial to generalisation and model adaptation; (ii) models transfer more easily across demographics but models trained on cross-domain data are more generalisable; (iii) some groups contribute more to generalisability than others; and (iv) dataset similarity is a signal of transferability.