StrucSum: Graph-Structured Reasoning for Long Document Extractive Summarization with LLMs

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Yuan, Haohan, Hong, Sukhwa, Zhang, Haopeng
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866911388112257024
author Yuan, Haohan
Hong, Sukhwa
Zhang, Haopeng
author_facet Yuan, Haohan
Hong, Sukhwa
Zhang, Haopeng
contents Large language models (LLMs) have shown strong performance in zero-shot summarization, but often struggle to model document structure and identify salient information in long texts. In this work, we introduce StrucSum, a training-free prompting framework that enhances LLM reasoning through sentence-level graph structures. StrucSum injects structural signals into prompts via three targeted strategies: Neighbor-Aware Prompting (NAP) for local context, Centrality-Aware Prompting (CAP) for importance estimation, and Centrality-Guided Masking (CGM) for efficient input reduction. Experiments on ArXiv, PubMed, and Multi-News demonstrate that StrucSum consistently improves both summary quality and factual consistency over unsupervised baselines and vanilla prompting. In particular, on ArXiv, it increases FactCC and SummaC by 19.2\% and 8.0\% points, demonstrating stronger alignment between summaries and source content. The ablation study shows that the combination of multiple strategies does not yield clear performance gains; therefore, structure-aware prompting with graph-based information represents a promising and underexplored direction for the advancement of zero-shot extractive summarization with LLMs. Our source code is publicly available.
format Preprint
id arxiv_https___arxiv_org_abs_2505_22950
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle StrucSum: Graph-Structured Reasoning for Long Document Extractive Summarization with LLMs
Yuan, Haohan
Hong, Sukhwa
Zhang, Haopeng
Computation and Language
Large language models (LLMs) have shown strong performance in zero-shot summarization, but often struggle to model document structure and identify salient information in long texts. In this work, we introduce StrucSum, a training-free prompting framework that enhances LLM reasoning through sentence-level graph structures. StrucSum injects structural signals into prompts via three targeted strategies: Neighbor-Aware Prompting (NAP) for local context, Centrality-Aware Prompting (CAP) for importance estimation, and Centrality-Guided Masking (CGM) for efficient input reduction. Experiments on ArXiv, PubMed, and Multi-News demonstrate that StrucSum consistently improves both summary quality and factual consistency over unsupervised baselines and vanilla prompting. In particular, on ArXiv, it increases FactCC and SummaC by 19.2\% and 8.0\% points, demonstrating stronger alignment between summaries and source content. The ablation study shows that the combination of multiple strategies does not yield clear performance gains; therefore, structure-aware prompting with graph-based information represents a promising and underexplored direction for the advancement of zero-shot extractive summarization with LLMs. Our source code is publicly available.
title StrucSum: Graph-Structured Reasoning for Long Document Extractive Summarization with LLMs
topic Computation and Language
url https://arxiv.org/abs/2505.22950