Commit
·
572a919
1
Parent(s):
b1adcaa
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,18 @@
|
|
| 1 |
---
|
| 2 |
license: cc-by-sa-4.0
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: cc-by-sa-4.0
|
| 3 |
---
|
| 4 |
+
# Utilising Weak Supervision to Create S3D: A Sarcasm Annotated Dataset
|
| 5 |
+
This is the repository for the S3D dataset published at EMNLP 2022. The dataset can help build sarcasm detection models.
|
| 6 |
+
|
| 7 |
+
# S3D Summary
|
| 8 |
+
The S3D dataset is our silver standard dataset of 100,000 tweets labelled for sarcasm using weak supervision by our **BERTweet-sarcasm-combined** model.
|
| 9 |
+
These tweets can be accessed by using the Twitter API so that they can be used for other experiments.
|
| 10 |
+
S3D contains 38879 tweets labelled as sarcastic, and 61211 tweets labelled as not being sarcastic.
|
| 11 |
+
# Data Fields
|
| 12 |
+
- Tweet ID: The ID of the labelled tweet
|
| 13 |
+
- Label: A label to denote if a given tweet is sarcastic
|
| 14 |
+
|
| 15 |
+
# Data Splits
|
| 16 |
+
- Train: 70,000
|
| 17 |
+
- Valid: 15,000
|
| 18 |
+
- Test: 15,000
|