---
license: mit
dataset_info:
  features:
  - name: text
    dtype: string
  - name: impoliteness
    dtype: int64
  - name: intolerance
    dtype: int64
  splits:
  - name: train
    num_bytes: 2169574.4014020115
    num_examples: 10498
  - name: test
    num_bytes: 542703.5985979884
    num_examples: 2626
  download_size: 1726706
  dataset_size: 2712278
task_categories:
- text-classification
language:
- en
---

## Overview
The TwitCivility dataset is specifically developed to classify political incivility, focusing on multidimensional aspects of impoliteness and intolerance.
Detailed methodologies are outlined in our [paper](https://arxiv.org/abs/2305.14964).


## Languages
All text is written in English.

## Dataset Structure
### Data Fields
We release TwitCivility as a data frame with the following fields: <br />
**text**: This field contains the text (after preprocessing and anonymization) of the tweet. <br />
**impoliteness**: A binary indicator (1 or 0) representing the presence of impoliteness in the text. A value of 1 signifies impoliteness, while 0 indicates non-impoliteness. <br />
**intolerance**: Similarly, this binary value denotes the presence of intolerance in the text, with 1 indicating intolerance and 0 signifying non-intolerance. <br />

## Citation Information
```
@misc{incivility2023,
      title={Detecting Multidimensional Political Incivility on Social Media}, 
      author={Sagi Pendzel and Nir Lotan and Alon Zoizner and Einat Minkov},
      year={2023},
      eprint={2305.14964},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}
```