Datasets:

dibo
/

hatevolution-vocabulary-expansion

license: cc-by-4.0
task_categories:
  - text-classification
language:
  - en
tags:
  - hate
  - speech
  - hate-speech
  - neologisms
  - social-media
size_categories:
  - n<1K

Dataset Info

The hatevolution-vocabulary-expansion dataset contains the data used for Experiment 2 in the paper Hatevolution: What Static Benchmarks Don't Tell Us (Di Bonaventura et al., 2025). It is built using the NeoBench dataset (Zheng et al., 2024), further annotated for hate speech detection.