File size: 672 Bytes
6c25ddb
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
## Intro

> TLDR: 
> This paper treats document-level argument extraction as a conditional generation problem by filling in event templates.


![Argument Extraction Model Overview](figures/model.png)

## Dataset

The WikiEvents datasets is under `data/wikievents`.

The RAMS dataset is for download [here](https://nlp.jhu.edu/rams/).

The ACE dataset is provided by LDC and unfortunately we cannot release it directly. See [this link](https://catalog.ldc.upenn.edu/LDC2006T06) for details.

## Code 

- v0.1 (April 12, 2021): Basic generation model for argument extraction. (This does not include the post-processing script.)



This page is currently under construction.