# Neural Abstractive Summarization with Structural Attention

arXiv: https://arxiv.org/abs/2004.09739 · HTML: https://mlsanity.com/paper/2004.09739/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 21 Apr 2020.

## Built on by (most cited first, 11 of 11)

- [See, Hear, Read: Leveraging Multimodality with Guided Attention for Abstractive Text Summarization](https://mlsanity.com/paper/2105.09601/) (2021)
- [A Survey on Legal Question Answering Systems](https://mlsanity.com/paper/2110.07333/) (2021)
- [AnswerSumm: A Manually-Curated Dataset and Pipeline for Answer Summarization](https://mlsanity.com/paper/2111.06474/) (2021)
- [Fusing Multimodal Signals on Hyper-complex Space for Extreme Abstractive Text Summarization (TL;DR) of Scientific Contents](https://mlsanity.com/paper/2306.13968/) (2023)
- [Summarization, Simplification, and Generation: The Case of Patents](https://mlsanity.com/paper/2104.14860/) (2021)
- [Towards Entity Alignment in the Open World: An Unsupervised Approach](https://mlsanity.com/paper/2101.10535/) (2021)
- [Multi-Perspective Abstractive Answer Summarization](https://mlsanity.com/paper/2104.08536/) (2021)
- [Aspect-oriented Consumer Health Answer Summarization](https://mlsanity.com/paper/2405.06295/) (2024)
- [Summarizing Community-based Question-Answer Pairs](https://mlsanity.com/paper/2211.09892/) (2022)
- [CISum: Learning Cross-modality Interaction to Enhance Multimodal Semantic Coverage for Multimodal Summarization](https://mlsanity.com/paper/2302.09934/) (2023)
- [YaleNLP @ PerAnsSumm 2025: Multi-Perspective Integration via Mixture-of-Agents for Enhanced Healthcare QA Summarization](https://mlsanity.com/paper/2504.03932/) (2025)

## Abstract (arXiv)

Attentional, RNN-based encoder-decoder architectures have achieved impressive performance on abstractive summarization of news articles. However, these methods fail to account for long term dependencies within the sentences of a document. This problem is exacerbated in multi-document summarization tasks such as summarizing the popular opinion in threads present in community question answering (CQA) websites such as Yahoo! Answers and Quora. These threads contain answers which often overlap or contradict each other. In this work, we present a hierarchical encoder based on structural attention to model such inter-sentence and inter-document dependencies. We set the popular pointer-generator architecture and some of the architectures derived from it as our baselines and show that they fail to generate good summaries in a multi-document setting. We further illustrate that our proposed model achieves significant improvement over the baselines in both single and multi-document summarization settings -- in the former setting, it beats the best baseline by 1.31 and 7.8 ROUGE-1 points on CNN and CQA datasets, respectively; in the latter setting, the performance is further improved by 1.6 ROUGE-1 points on the CQA dataset.