Enhancing Long Document Long Form Summarisation with Self-Planning

TL;DR

This paper presents a new highlight-guided summarisation method for long documents that uses sentence-level content plans to enhance factual accuracy and detail preservation, outperforming existing approaches.

Contribution

It introduces a self-planning framework with end-to-end and two-stage variants, demonstrating improved factual consistency and relevance in long-form summarisation tasks.

Findings

Two-stage pipeline outperforms end-to-end on dense documents.

Achieves 4.1 ROUGE-L point improvement on GovReport.

35% gains in SummaC factual consistency scores.

Abstract

We introduce a novel approach for long context summarisation, highlight-guided generation, that leverages sentence-level information as a content plan to improve the traceability and faithfulness of generated summaries. Our framework applies self-planning methods to identify important content and then generates a summary conditioned on the plan. We explore both an end-to-end and two-stage variants of the approach, finding that the two-stage pipeline performs better on long and information-dense documents. Experiments on long-form summarisation datasets demonstrate that our method consistently improves factual consistency while preserving relevance and overall quality. On GovReport, our best approach has improved ROUGE-L by 4.1 points and achieves about 35% gains in SummaC scores. Qualitative analysis shows that highlight-guided summarisation helps preserve important details, leading to more accurate and insightful summaries across domains.