Update README.md

README.md

@@ -1,25 +1,17 @@
 ---
 license: cc-by-sa-4.0
 ---
-# Utilising Weak Supervision to Create a Sarcasm Annotated Dataset: S3D
-
+# 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.
 
 # SAD
-
-The SAD dataset is our gold standard dataset of tweets labelled for sarcasm. These tweets were scraped observing a #sarcasm hashtag and then manually annotated by three annotators. There are a total of 2340 tweets, 1170 are sarcastic and 1170 are denoted otherwise. 
+The SAD dataset is our gold standard dataset of tweets labelled for sarcasm. These tweets were scraped by observing a '#sarcasm' hashtag and then manually annotated by three annotators. There are a total of 1170 pairs of a sarcastic and non-sarcastic tweets which were both posted by the same user, resulting in a total of 2340 tweets annotated for sarcasm. 
 
 # Data Fields
-
 - Tweet ID: The ID of the labelled tweet
 - Label: A label to denote if a given tweet is sarcastic
 
 # Data Splits
-
 - Train: 1638
 - Valid: 351
-- Test: 351
-
-# Languages
-
-- English
+- Test: 351