Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
csv
Languages:
Korean
Size:
10K - 100K
ArXiv:
License:
metadata
annotations_creators:
- crowdsourced
- crowd-generated
language_creators:
- found
language:
- ko
license:
- cc-by-sa-4.0
multilinguality:
- monolingual
paperswithcode_id: apeach
pretty_name: APEACH
size_categories:
- 1K<n<10K
source_datasets:
- original
task_categories:
- text-classification
task_ids:
- binary-classification
Dataset for project: kor_hate_eval(APEACH)
Sample Code
Dataset Descritpion
Korean Hate Speech Evaluation Datasets : trained with BEEP! and evaluate with APEACH
- Repository: Korean HateSpeech Evaluation Dataset
- Paper: APEACH: Attacking Pejorative Expressions with Analysis on Crowd-Generated Hate Speech Evaluation Datasets
- Point of Contact: Kichang Yang
Languages
ko-KR
Dataset Structure
Data Instances
A sample from this dataset looks as follows:
{'text': ['(ํ์ฌ ํธํ
์ฃผ์ธ ์ฌ์ ) ์18 ๋ ๋ง๋ฅธํ๋์ ๋ ๋ฒผ๋ฝ๋ง๊ณ ํธํ
๋งํ๊ฒ์๊ฒผ๋๋ฐ ๋๊ตฐ ๊ณ์ ์ถ๋ชจ๋ฐ๋ค....',
'....ํ๊ตญ์ ์ธ ๋ฏธ์ธ์ ๋ํ์ ์ธ ๋ถ...๋๋ฌด๋ ๊ณฑ๊ณ ์๋ฆ๋ค์ด๋ชจ์ต...๊ทธ๋ชจ์ต๋ค์ ์ฌํ์ ๋ฏธ์ฒ ์์ง๋ชปํ๋ค์ใ
'],
'class': ['Spoiled', 'Default']}
Dataset Fields
The dataset has the following fields (also called "features"):
{
"text": "Value(dtype='string', id=None)",
"class": "ClassLabel(num_classes=2, names=['Default', 'Spoiled'], id=None)"
}
Dataset Splits
This dataset is split into a train and validation split. The split sizes are as follow:
Split name | Num samples |
---|---|
train (binarized BEEP!) | 7896 |
valid (APEACH) | 3770 |
Citation
@article{yang2022apeach,
title={APEACH: Attacking Pejorative Expressions with Analysis on Crowd-Generated Hate Speech Evaluation Datasets},
author={Yang, Kichang and Jang, Wonjun and Cho, Won Ik},
journal={arXiv preprint arXiv:2202.12459},
year={2022}
}