Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,28 @@
|
|
| 1 |
---
|
| 2 |
license: mit
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: mit
|
| 3 |
+
task_categories:
|
| 4 |
+
- summarization
|
| 5 |
+
language:
|
| 6 |
+
- en
|
| 7 |
+
pretty_name: aclsum
|
| 8 |
+
size_categories:
|
| 9 |
+
- n<1K
|
| 10 |
---
|
| 11 |
+
|
| 12 |
+
# ACLSum: A New Dataset for Aspect-based Summarization of Scientific Publications
|
| 13 |
+
|
| 14 |
+
This repository contains data for our paper "ACLSum: A New Dataset for Aspect-based Summarization of Scientific Publications" and a small
|
| 15 |
+
utility class to work with it.
|
| 16 |
+
|
| 17 |
+
|
| 18 |
+
## HuggingFace datasets
|
| 19 |
+
|
| 20 |
+
You can also use Huggin Face datasets to load ACLSum ([dataset link](https://huggingface.co/datasets/sobamchan/aclsum)).
|
| 21 |
+
This would be convenient if you want to train transformer models using our dataset.
|
| 22 |
+
|
| 23 |
+
Just do,
|
| 24 |
+
|
| 25 |
+
```py
|
| 26 |
+
from datasets import load_dataset
|
| 27 |
+
dataset = load_dataset("sobamchan/aclsum")
|
| 28 |
+
```
|