fkdosilovic commited on
Commit
cb59848
1 Parent(s): 193d9b3

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +24 -4
README.md CHANGED
@@ -20,13 +20,33 @@ task_ids:
20
  - multi-class-classification
21
  ---
22
 
23
- This is a [DocEE dataset](https://github.com/tongmeihan1995/docee) adapted for easy use with HuggingFace ecosystem.
24
 
25
- Dataset is described in [DocEE: A Large-Scale and Fine-grained Benchmark for Document-level Event Extraction](https://aclanthology.org/2022.naacl-main.291/) and can be downloaded from [here](https://github.com/tongmeihan1995/docee).
26
 
27
- Train split contains 21949 documents while test split contains 5536 documents. In total, dataset contains 27485 documents classified into 59 event types.
 
 
28
 
29
- ## Differences from the original split(s)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
30
 
31
  Originally, the dataset is split into three splits: train, validation and test. For the purposes of this repository, original splits were joined back together and divided into train and test splits while making sure that splits were stratified across document sources (news and wiki) and event types.
32
 
 
20
  - multi-class-classification
21
  ---
22
 
23
+ # Dataset Card for DocEE Dataset
24
 
25
+ ## Dataset Description
26
 
27
+ - **Homepage:**
28
+ - **Repository:** [DocEE Dataset repository](https://github.com/tongmeihan1995/docee)
29
+ - **Paper:** [DocEE: A Large-Scale and Fine-grained Benchmark for Document-level Event Extraction](https://aclanthology.org/2022.naacl-main.291/)
30
 
31
+ ### Dataset Summary
32
+
33
+ DocEE dataset is an English-language dataset containing more than 27k news and Wikipedia articles. Dataset is primarily annotated and collected for large-scale document-level event extraction.
34
+
35
+ ### Data Fields
36
+
37
+ - `title`: TODO
38
+ - `text`: TODO
39
+ - `event_type`: TODO
40
+ - `date`: TODO
41
+ - `metadata`: TODO
42
+
43
+ **Note: this repo contains only event detection portion of the dataset.**
44
+
45
+ ### Data Splits
46
+
47
+ The dataset has 2 splits: _train_ and _test_. Train split contains 21949 documents while test split contains 5536 documents. In total, dataset contains 27485 documents classified into 59 event types.
48
+
49
+ #### Differences from the original split(s)
50
 
51
  Originally, the dataset is split into three splits: train, validation and test. For the purposes of this repository, original splits were joined back together and divided into train and test splits while making sure that splits were stratified across document sources (news and wiki) and event types.
52