liyucheng commited on
Commit
6e7da56
·
1 Parent(s): 4367f3d

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +20 -2
README.md CHANGED
@@ -48,6 +48,24 @@ dataset_info:
48
  download_size: 1038853
49
  dataset_size: 7299373
50
  ---
51
- # Dataset Card for "EPIC_Irony"
52
 
53
- [More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
48
  download_size: 1038853
49
  dataset_size: 7299373
50
  ---
51
+ # EPIC_Irony
52
 
53
+ - paper: [EPIC: Multi-Perspective Annotation of a Corpus of Irony](https://assets.amazon.science/40/b4/0f6ec06a4a33a44485de1b2b57c7/epic-multi-perspective-annotation-of-a-corpus-of-irony.pdf) at ACL 2023
54
+
55
+ Key features:
56
+
57
+ - EPIC (English Perspectivist Irony Corpus) is the first annotated corpus specifically created for irony analysis based on the principles of data perspectivism.
58
+ - The corpus contains short social media conversations in five regional varieties of English, annotated by contributors from five countries corresponding to those varieties.
59
+ - The analysis of the resource considers the perspectives of the annotators in terms of origin, age, and gender, and the relationship between these dimensions, irony, and the topics of conversation.
60
+ - To validate EPIC, perspective-aware models are created that encode the perspectives of annotators based on their demographic characteristics.
61
+ - The performance of perspectivist models confirms that different annotators induce very different models.
62
+ - In classifying ironic and non-ironic texts, perspectivist models prove to be generally more confident than non-perspectivist ones.
63
+ - Perspectivist models tend to more accurately detect ironic language, indicating their ability to capture the different perceptions of irony.
64
+ - The models reveal interesting insights about the variation in the perception of irony among different groups of annotators, such as among different generations and nationalities.
65
+ - The EPIC corpus provides a useful resource for training perspective-aware models for irony detection, and highlights the influence of demographic factors on the perception and understanding of irony.
66
+
67
+ Metadata in Creative Language Toolkit ([CLTK](https://github.com/liyucheng09/cltk))
68
+ - CL Type: Irony
69
+ - Task Type: detection
70
+ - Size: 14k
71
+ - Created time: 2023