Capturing Relations between Scientific Papers: An Abstractive Model for Related Work Section Generation
This repository contains dataset for the ACL 2021 paper Capturing Relations between Scientific Papers: An Abstractive Model for Related Work Section Generation.
S2ORC corpus consists of 126,655 reference-relatedwork training pairs, 5000 validation and test pairs. Delve corpus consists of 72,927 reference-relatedwork training pairs, 3000 validation and test pairs.
Signed the following copyright announcement with your name and organization. Then complete the form online(shorturl.at/dpMY3) and mail to xy-chen#pku.edu.cn ('#'->'@'), we will send you the corpus by e-mail when approved.
The original copyright belongs to the source owner.
The copyright of annotation belongs to our group, and they are free to the public.
The dataset is only for research purposes. Without permission, it may not be used for any commercial purposes and distributed to others.
We appreciate your citation if you find our dataset is beneficial.
@inproceedings{chen2021related,
title={Capturing Relations between Scientific Papers: An Abstractive Model for Related Work Section Generation},
author={Xiuying Chen, Hind Alamro, Mingzhe Li, Shen Gao, Xiangliang Zhang, Dongyan Zhao, Rui Yan},
booktitle = {ACL},
year = {2021}
}