cdt

Stay organized with collections Save and categorize content based on your preferences.

References:

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:cdt')
  • Description:
The Cyberbullying Detection task was part of 2019 edition of PolEval competition. The goal is to predict if a given Twitter message contains a cyberbullying (harmful) content.
  • License: BSD 3-Clause
  • Version: 1.1.0
  • Splits:
Split Examples
'test' 1000
'train' 10041
  • Features:
{
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "target": {
        "num_classes": 2,
        "names": [
            "0",
            "1"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}