|
--- |
|
language: |
|
- en |
|
task_categories: |
|
- text-classification |
|
--- |
|
# AutoNLP Dataset for project: tweet-sentiment |
|
|
|
## Table of content |
|
- [Dataset Description](#dataset-description) |
|
- [Languages](#languages) |
|
- [Dataset Structure](#dataset-structure) |
|
- [Data Instances](#data-instances) |
|
- [Data Fields](#data-fields) |
|
- [Data Splits](#data-splits) |
|
|
|
## 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: |
|
|
|
```json |
|
[ |
|
{ |
|
"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"): |
|
|
|
```json |
|
{ |
|
"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 | |
|
|
