# Dealing with Annotator Disagreement in Hate Speech Classification

arXiv: https://arxiv.org/abs/2502.08266 · HTML: https://mlsanity.com/paper/2502.08266/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Safety alignment, refusal and guardrails, one of 12 directions in LLM safety and security.
- Builds on 14 of its 52 references that are in the map.
- First posted on arXiv 12 Feb 2025.

## Placed in

- LLM safety and security / [Safety alignment, refusal and guardrails](https://mlsanity.com/safety/safety-alignment-refusal-and-guardrails/)

## Builds on (most cited first, 14 of 14)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [MultiVerS: Improving scientific claim verification with weak supervision and full-document context](https://mlsanity.com/paper/2112.01640/) (2021)
- [On Releasing Annotator-Level Labels and Information in Datasets](https://mlsanity.com/paper/2110.05699/) (2021)
- [When the Majority is Wrong: Modeling Annotator Disagreement for Subjective Tasks](https://mlsanity.com/paper/2305.06626/) (2023)
- [Everyone's Voice Matters: Quantifying Annotation Disagreement Using Demographic Information](https://mlsanity.com/paper/2301.05036/) (2023)
- [SemEval-2023 Task 11: Learning With Disagreements (LeWiDi)](https://mlsanity.com/paper/2304.14803/) (2023)
- [You Are What You Annotate: Towards Better Models through Annotator Representations](https://mlsanity.com/paper/2305.14663/) (2023)
- [NER-BERT: A Pre-trained Model for Low-Resource Entity Tagging](https://mlsanity.com/paper/2112.00405/) (2021)
- [Let's agree to disagree: learning highly debatable multirater labelling](https://mlsanity.com/paper/1909.01891/) (2019)
- [The Origin and Value of Disagreement Among Data Labelers: A Case Study of the Individual Difference in Hate Speech Annotation](https://mlsanity.com/paper/2112.04030/) (2021)
- [Hate Speech Classifiers Learn Human-Like Social Stereotypes](https://mlsanity.com/paper/2110.14839/) (2021)
- [Annotating for Hate Speech: The MaNeCo Corpus and Some Input from Critical Discourse Analysis](https://mlsanity.com/paper/2008.06222/) (2020)
- [Investigating Annotator Bias in Large Language Models for Hate Speech Detection](https://mlsanity.com/paper/2406.11109/) (2024)
- [Factoring Hate Speech: A New Annotation Framework to Study Hate Speech in Social Media](https://mlsanity.com/paper/2311.03969/) (2023)

## Built on by (most cited first, 3 of 3)

- [Rethinking Failure Attribution in Multi-Agent Systems: A Multi-Perspective Benchmark and Evaluation](https://mlsanity.com/paper/2603.25001/) (2026)
- [Can LLMs Evaluate What They Cannot Annotate? Revisiting LLM Reliability in Hate Speech Detection](https://mlsanity.com/paper/2512.09662/) (2025)
- [Hate Speech Detection in Turkish and Arabic: A Comprehensive Study](https://mlsanity.com/paper/2607.00143/) (2026)

## Abstract (arXiv)

Hate speech detection is a crucial task, especially on social media where harmful content can spread quickly. Collecting social media content (tweets etc.) to train machine learning models is easy, but detecting and categorizing hate speech can be difficult due to the inherently subjective nature. This subjectivity leads to frequent disagreement among annotators, particularly for subtle or borderline content. Traditional approaches either discard non-consensus samples or force a ''gold standard'' through expert adjudication, ignoring valuable information about uncertainty and diverse human perspectives. We examine the largely overlooked problem of annotator disagreement in hate speech classification and evaluate a range of aggregation methods, including majority voting, ordinal strategies (minimum, maximum, and mean), and analyze their impact across binary, 4-class, and 6-class classification tasks. In addition, we leverage annotators' perceived hate speech strength scores to explore regression-based and hybrid modeling approaches. Among others, we show that filtering non-consensus samples results in over-optimistic results and that the perceived strength provides a complementary signal that enhance classification performance. Finally, we establish new state-of-the-art results for hate speech detection in Turkish tweets, and demonstrate that annotator disagreement, when properly modeled, is a valuable resource for building more robust and reliable systems.