README.md

metadata

annotations_creators:
  - expert-generated
language_creators:
  - found
language:
  - da
license:
  - cc-by-4.0
multilinguality:
  - monolingual
pretty_name: DKHate
size_categories:
  - 1K<n<10K
source_datasets:
  - original
task_categories:
  - text-classification
task_ids:
  - hate-speech-detection
extra_gated_prompt: >-
  Content warning: This dataset contains harmful text (abusive language, hate
  speech).
paperswithcode_id: dkhate

Dataset Card for DKHate

Table of Contents

• Table of Contents
• Dataset Description
  • Dataset Summary
  • Languages
• Dataset Structure
  • Data Instances
  • Data Fields
  • Data Splits
• Additional Information
  • Dataset Curators
  • Licensing Information
  • Citation Information
  • Contributions

Dataset Description

• Homepage: https://stromberg.ai/publication/offensivelanguageandhatespeechdetectionfordanish/
• Repository: https://github.com/StrombergNLP/dkhate
• Paper: https://https://aclanthology.org/2020.lrec-1.430/, https://arxiv.org/abs/1908.04531
• Direct Download: https://figshare.com/articles/dataset/Danish_Hate_Speech_Abusive_Language_data/12220805
• Point of Contact: Leon Derczynski

Dataset Summary

This dataset consists of anonymised Danish Twitter data that has been annotated for hate speech. All credits go to the authors of the following paper, who created the dataset:

Offensive Language and Hate Speech Detection for Danish (Sigurbergsson & Derczynski, LREC 2020)

Supported Tasks and Leaderboards

This dataset is suitable for hate speech detection.

• PwC leaderboard for Task A: Hate Speech Detection on DKhate

Languages

This dataset is in Danish.

Dataset Structure

Data Instances

Every entry in the dataset has a tweet and an associated label.

Data Fields

An entry in the dataset consists of the following fields:

• text (str): The tweet content.
• label (str): The label of the text. Can be either "OFF" or "NOT", being offensive and not offensive, respectively.

Data Splits

A train and test split is available, which are identical to the original splits. There are 2,960 tweets in the training split and 329 in the test split.

Additional Information

Dataset Curators

The curation of the dataset is solely due to the authors of the original paper: Gudbjartur Ingi Sigurbergsson and Leon Derczynski.

Licensing Information

The dataset is released under the CC BY 4.0 license.

Citation Information

@inproceedings{sigurbergsson2020offensive,
  title={Offensive Language and Hate Speech Detection for Danish},
  author={Sigurbergsson, Gudbjartur Ingi and Derczynski, Leon},
  booktitle={Proceedings of the 12th Language Resources and Evaluation Conference},
  pages={3498--3508},
  year={2020}
}

Contributions

Thanks to @saattrupdan for adding this dataset to the Hugging Face Hub.