README.md

metadata

dataset_info:
  features:
    - name: text
      dtype: string
    - name: label
      dtype:
        class_label:
          names:
            '0': NOT
            '1': 'OFF'
  splits:
    - name: train
      num_bytes: 4133465
      num_examples: 31756
    - name: test
      num_bytes: 467184
      num_examples: 3528
  download_size: 3181241
  dataset_size: 4600649
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
license: cc
task_categories:
  - text-classification
language:
  - tr
size_categories:
  - 10K<n<100K

OffensEval-TR 2020 (Mirror)

Note: This is a mirror of the original dataset at coltekin/offenseval2020_tr. This copy was created due to HuggingFace's deprecation of trust_remote_code in datasets library version 4.0.0 (release notes), which makes the original dataset inaccessible with newer versions of the datasets library. This copy will be removed upon request from the original authors or if the original repository is updated to work without remote code.

Table of Contents

• Dataset Description
  • Dataset Summary
  • Supported Tasks and Leaderboards
  • Languages
• Dataset Structure
  • Data Instances
  • Data Fields
  • Data Splits
• Additional Information
  • Dataset Curators
  • Licensing Information
  • Citation Information
  • Contributions

Dataset Description

• Original Dataset: coltekin/offenseval2020_tr
• Homepage: offensive-turkish
• Paper: A Corpus of Turkish Offensive Language on Social Media
• Point of Contact: Çağrı Çöltekin

Dataset Summary

Twitter user mentions were substituted by @USER and URLs have been substitute by URL.

Each instance contains up to 1 labels corresponding to one of the following sub-task:

• Sub-task A: Offensive language identification;

Supported Tasks and Leaderboards

The dataset was published on this paper.

Languages

The dataset is based on Turkish.

Dataset Structure

Data Instances

A binary dataset with with (NOT) Not Offensive and (OFF) Offensive tweets.

Data Fields

The column names in the file are the following:

text label

The labels used in the annotation are listed below.

Task and Labels

(A) Sub-task A: Offensive language identification

• (NOT) Not Offensive - This post does not contain offense or profanity.
• (OFF) Offensive - This post contains offensive language or a targeted (veiled or direct) offense

In our annotation, we label a post as offensive (OFF) if it contains any form of non-acceptable language (profanity) or a targeted offense, which can be veiled or direct.

Data Splits

train	test
31756	3528

Additional Information

Dataset Curators

This dataset was originally curated by Çağrı Çöltekin. This mirror was created to maintain accessibility following HuggingFace's deprecation of remote code execution in the datasets library.

Licensing Information

The annotations are distributed under the terms of Creative Commons Attribution License (CC-BY 2.0). Please cite the following paper if you use this resource.

Citation Information

Please cite the original paper:

@inproceedings{coltekin2020lrec,
 author  = {\c{C}\"{o}ltekin, \c{C}a\u{g}r{\i}},
 year  = {2020},
 title  = {A Corpus of Turkish Offensive Language on Social Media},
 booktitle  = {Proceedings of The 12th Language Resources and Evaluation Conference},
 pages  = {6174--6184},
 address  = {Marseille, France},
 url  = {https://www.aclweb.org/anthology/2020.lrec-1.758},
}

Contributions

Thanks to @yavuzKomecoglu for adding the original dataset to HuggingFace.

---

Mirror Information

Original Dataset: coltekin/offenseval2020_tr

Reason for Mirror: HuggingFace deprecated trust_remote_code in datasets library v4.0.0, making the original dataset inaccessible with newer versions of the library.