Citance-Contextualized Summarization of Scientific Papers

TL;DR

This paper introduces a novel method for generating contextualized summaries of scientific papers that focus on the relationship between the citing sentence and the cited work, enhancing understanding of citation context.

Contribution

It presents a new approach that models citances, retrieves relevant passages from cited papers, and generates tailored summaries, addressing limitations of traditional abstract-based summarization.

Findings

Achieved effective citation-contextualized summaries on a large dataset

Demonstrated improved relevance over standard summarization methods

Utilized a new dataset with 540K papers and 4.6M citances

Abstract

Current approaches to automatic summarization of scientific papers generate informative summaries in the form of abstracts. However, abstracts are not intended to show the relationship between a paper and the references cited in it. We propose a new contextualized summarization approach that can generate an informative summary conditioned on a given sentence containing the citation of a reference (a so-called "citance"). This summary outlines the content of the cited paper relevant to the citation location. Thus, our approach extracts and models the citances of a paper, retrieves relevant passages from cited papers, and generates abstractive summaries tailored to each citance. We evaluate our approach using $\textbf{Webis-Context-SciSumm-2023}$, a new dataset containing 540K~computer science papers and 4.6M~citances therein.