# Do We Really Need That Many Parameters In Transformer For Extractive Summarization? Discourse Can Help !

arXiv: https://arxiv.org/abs/2012.02144 · HTML: https://mlsanity.com/paper/2012.02144/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 3 Dec 2020.

## Built on by (most cited first, 10 of 10)

- [Incorporating Distributions of Discourse Structure for Long Document Abstractive Summarization](https://mlsanity.com/paper/2305.16784/) (2023)
- [GCDT: A Chinese RST Treebank for Multigenre and Multilingual Discourse Parsing](https://mlsanity.com/paper/2210.10449/) (2022)
- [Predicting Discourse Trees from Transformer-based Neural Summarizers](https://mlsanity.com/paper/2104.07058/) (2021)
- [Demoting the Lead Bias in News Summarization via Alternating Adversarial Learning](https://mlsanity.com/paper/2105.14241/) (2021)
- [Human Guided Exploitation of Interpretable Attention Patterns in Summarization and Topic Segmentation](https://mlsanity.com/paper/2112.05364/) (2021)
- [EDU-level Extractive Summarization with Varying Summary Lengths](https://mlsanity.com/paper/2210.04029/) (2022)
- [Discourse Structure Extraction from Pre-Trained and Fine-Tuned Language Models in Dialogues](https://mlsanity.com/paper/2302.05895/) (2023)
- [Unsupervised Mutual Learning of Discourse Parsing and Topic Segmentation in Dialogue](https://mlsanity.com/paper/2405.19799/) (2024)
- [W-RST: Towards a Weighted RST-style Discourse Framework](https://mlsanity.com/paper/2106.02658/) (2021)
- [Towards Domain-Independent Supervised Discourse Parsing Through Gradient Boosting](https://mlsanity.com/paper/2210.09565/) (2022)

## Abstract (arXiv)

The multi-head self-attention of popular transformer models is widely used within Natural Language Processing (NLP), including for the task of extractive summarization. With the goal of analyzing and pruning the parameter-heavy self-attention mechanism, there are multiple approaches proposing more parameter-light self-attention alternatives. In this paper, we present a novel parameter-lean self-attention mechanism using discourse priors. Our new tree self-attention is based on document-level discourse information, extending the recently proposed "Synthesizer" framework with another lightweight alternative. We show empirical results that our tree self-attention approach achieves competitive ROUGE-scores on the task of extractive summarization. When compared to the original single-head transformer model, the tree attention approach reaches similar performance on both, EDU and sentence level, despite the significant reduction of parameters in the attention component. We further significantly outperform the 8-head transformer model on sentence level when applying a more balanced hyper-parameter setting, requiring an order of magnitude less parameters.