Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
csv
Languages:
English
Size:
10K - 100K
License:
annotations_creators: | |
- Jordan Painter, Diptesh Kanojia | |
language: | |
- en | |
license: | |
- cc-by-sa-4.0 | |
multilinguality: | |
- monolingual | |
pretty_name: 'Utilising Weak Supervision to create S3D: A Sarcasm Annotated Dataset' | |
size_categories: | |
- 100K<n<1M | |
source_datasets: | |
- original | |
task_categories: | |
- text-classification | |
## Table of Contents | |
- [Dataset Description](#dataset-description) | |
- | |
# Utilising Weak Supervision to Create S3D: A Sarcasm Annotated Dataset | |
This is the repository for the S3D dataset published at EMNLP 2022. The dataset can help build sarcasm detection models. | |
# S3D-v2 Summary | |
The S3D-v2 dataset is our silver standard dataset of 100,000 tweets labelled for sarcasm using weak supervision by a majority voting system of fine-tuned sarcasm detection models. The models used are | |
our [roberta-large-finetuned-SARC-combined-DS](https://huggingface.co/surrey-nlp/roberta-large-finetuned-SARC-combined-DS), [bertweet-base-finetuned-SARC-DS](https://huggingface.co/surrey-nlp/bertweet-base-finetuned-SARC-DS) | |
and [bertweet-base-finetuned-SARC-combined-DS](https://huggingface.co/surrey-nlp/bertweet-base-finetuned-SARC-combined-DS) models. | |
S3D contains 13016 tweets labelled as sarcastic, and 86904 tweets labelled as not being sarcastic. | |
# Data Fields | |
- Text: The preprocessed tweet | |
- Label: A label to denote if a given tweet is sarcastic | |
# Data Splits | |
- Train: 70,000 | |
- Valid: 15,000 | |
- Test: 15,000 |