Capturing Relations between Scientific Papers: An Abstractive Model for Related Work Section Generation

Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, pages 6068–6077 August 1–6, 2021. ©2021 Association for Computational Linguistics 6068 Capturing Relations between Scientific Papers: An Abstractive Model for Related Work Section Generation Xiuying Chen123, Hind Alamro34, Mingzhe Li12, Shen Gao1, Xiangliang Zhang3∗, Dongyan Zhao125∗, Rui Yan6 1Wangxuan Institute of Computer Technology, Peking University, Beijing,China 2Center for Data Science, Peking University, Beijing, China 3King Abdullah University of Science and Technology, Thuwal, Saudi Arabia 4Computer Science Department, Umm Al-Qura University, Makkah, Saudi Arabia 5State Key Laboratory of Media Convergence Production Technology and Systems 6Gaoling School of Artificial Intelligence, Renmin University of China xy-chen@pku.edu.cn Abstract Given a set of related publications, related work section generation aims to provide re- searchers with an overview of the specific re- search area by summarizing these works and introducing them in a logical order. Most of existing related work section generation mod- els follow the inflexible extractive style, which directly extract sentences from multiple origi- nal papers to form a related work discussion. Hence, in this paper, we propose a Relation- aware Related work Generator (RRG), which generates an abstractive related work section from multiple scientific papers in the same re- search area. Concretely, we propose a relation- aware multi-document encoder that relates one document to another according to their con- tent dependency in a relation graph. The rela- tion graph and the document representation in- teract and are refined iteratively, complement- ing each other in the training process. We also contribute two public datasets composed of related work sections and their correspond- ing papers1. Extensive experiments on the two datasets show that the proposed model brings substantial improvements over several strong baselines. We hope that this work will pro- mote advances in related work section gener- ation task. 1 Introduction The related work section generation task aims to automatically generate a summary of the most rele- vant works in a specific research area, which can help researchers to familiarize themselves with the state of the art in the field. Several methods (Hoang and Kan, 2010; Hu and Wan, 2014; Chen and Zhuge, 2019) have been proposed to study how to obtain the related work section automatically by ∗Corresponding author. 1https://github.com/iriscxy/ relatedworkgeneration Extractive Related Work: We find that CRISPR/Cas9 can robustly and specifically reduce the expression of these mi- croRNAs up to 96% [1]. We find that miRNA knockdown phenotypes caused by CRISPR/Cas9 transient editing can be stably maintained in both in vitro and in vivo models for a long term (up to 30 days) [2]. Although genome edit- ing using the CRISPR-Cas system is highly efficient in human cell lines, CRISPR-Cas genome editing in primary human cells is more challenging [3]. Abstractive Related Work: Recently, [1] showed that CRISPER-Cas9 targeted miRNA-17, miRNA-200c and miRNA-141, repressed their activity in human colon can- cer cell lines HCT116 and HT-29. Furthermore, in vivo targeting was effective for at least a month [2]. However, off-target mutagenesis …

Content truncated to 500 words