Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
csv
Languages:
English
Size:
10K - 100K
License:
Commit
·
7695b6b
1
Parent(s):
9b2610b
Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,39 @@
|
|
1 |
---
|
2 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
+
annotations_creators:
|
3 |
+
- Jordan Painter, Diptesh Kanojia
|
4 |
+
language:
|
5 |
+
- en
|
6 |
+
license:
|
7 |
+
- cc-by-sa-4.0
|
8 |
+
multilinguality:
|
9 |
+
- monolingual
|
10 |
+
pretty_name: 'Utilising Weak Supervision to create S3D: A Sarcasm Annotated Dataset'
|
11 |
+
size_categories:
|
12 |
+
- 100K<n<1M
|
13 |
+
source_datasets:
|
14 |
+
- original
|
15 |
+
task_categories:
|
16 |
+
- text-classification
|
17 |
---
|
18 |
+
|
19 |
+
## Table of Contents
|
20 |
+
- [Dataset Description](#dataset-description)
|
21 |
+
|
22 |
+
-
|
23 |
+
# Utilising Weak Supervision to Create S3D: A Sarcasm Annotated Dataset
|
24 |
+
This is the repository for the S3D dataset published at EMNLP 2022. The dataset can help build sarcasm detection models.
|
25 |
+
|
26 |
+
# S3D-v2 Summary
|
27 |
+
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
|
28 |
+
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)
|
29 |
+
and [bertweet-base-finetuned-SARC-combined-DS](https://huggingface.co/surrey-nlp/bertweet-base-finetuned-SARC-combined-DS).
|
30 |
+
|
31 |
+
S3D contains 13016 tweets labelled as sarcastic, and 86904 tweets labelled as not being sarcastic.
|
32 |
+
# Data Fields
|
33 |
+
|
34 |
+
- Label: A label to denote if a given tweet is sarcastic
|
35 |
+
|
36 |
+
# Data Splits
|
37 |
+
- Train: 70,000
|
38 |
+
- Valid: 15,000
|
39 |
+
- Test: 15,000
|