albertvillanova's picture
Fix language and license tag names (#1)
80ce985
metadata
language:
  - en
task_categories:
  - text-classification

AutoNLP Dataset for project: tweet-sentiment

Table of content

Dataset Descritpion

This dataset has been automatically processed by AutoNLP for project tweet-sentiment.

Languages

The BCP-47 code for the dataset's language is en.

Dataset Structure

Data Instances

A sample from this dataset looks as follows:

[
  {
    "text": "I am going to see how long I can do this for.",
    "target": 8
  },
  {
    "text": "@anitabora yeah, right. What if our politicians start using uploading their pics, lots of inside sto[...]",
    "target": 8
  }
]

Dataset Fields

The dataset has the following fields (also called "features"):

{
  "text": "Value(dtype='string', id=None)",
  "target": "ClassLabel(num_classes=13, names=['anger', 'boredom', 'empty', 'enthusiasm', 'fun', 'happiness', 'hate', 'love', 'neutral', 'relief', 'sadness', 'surprise', 'worry'], 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 31995
valid 8005