README.md

metadata

annotations_creators:
  - crowdsourced
language:
  - en
language_creators:
  - found
license: []
multilinguality:
  - monolingual
pretty_name: RuyuanWan/SBIC_Disagreement
size_categories: []
source_datasets:
  - extended|social_bias_frames
tags: []
task_categories:
  - text-classification
task_ids: []

This dataset is processed version of Social Bias Inference Corpus(SBIC) dataset including text, annotator's demographics and the annotation disagreement labels.

Paper: Everyone's Voice Matters: Quantifying Annotation Disagreement Using Demographic Information
Authors: Ruyuan Wan, Jaehyung Kim, Dongyeop Kang
Github repo: https://github.com/minnesotanlp/Quantifying-Annotation-Disagreement