Enhancing Abstractive Summarization of Scientific Papers Using Structure Information

TL;DR

This paper introduces a two-stage framework for scientific paper summarization that leverages structural information recognition and Longformer-based context modeling to produce more comprehensive and balanced summaries.

Contribution

It presents a novel two-stage approach that automatically recognizes structural components and employs Longformer for improved abstractive summarization of scientific papers.

Findings

Outperforms existing baselines on domain-specific datasets

Effectively captures structural information for balanced summaries

Demonstrates robustness across different scientific disciplines

Abstract

Abstractive summarization of scientific papers has always been a research focus, yet existing methods face two main challenges. First, most summarization models rely on Encoder-Decoder architectures that treat papers as sequences of words, thus fail to fully capture the structured information inherent in scientific papers. Second, existing research often use keyword mapping or feature engineering to identify the structural information, but these methods struggle with the structural flexibility of scientific papers and lack robustness across different disciplines. To address these challenges, we propose a two-stage abstractive summarization framework that leverages automatic recognition of structural functions within scientific papers. In the first stage, we standardize chapter titles from numerous scientific papers and construct a large-scale dataset for structural function recognition. A classifier is then trained to automatically identify the key structural components (e.g., Background, Methods, Results, Discussion), which provides a foundation for generating more balanced summaries. In the second stage, we employ Longformer to capture rich contextual relationships across sections and generating context-aware summaries. Experiments conducted on two domain-specific scientific paper summarization datasets demonstrate that our method outperforms advanced baselines, and generates more comprehensive summaries. The code and dataset can be accessed at https://github.com/tongbao96/code-for-SFR-AS.