# Neural Generative Rhetorical Structure Parsing

arXiv: https://arxiv.org/abs/1909.11049 · HTML: https://mlsanity.com/paper/1909.11049/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 24 Sep 2019.

## Built on by (most cited first, 5 of 5)

- [Deep Discourse Analysis for Generating Personalized Feedback in Intelligent Tutor Systems](https://mlsanity.com/paper/2103.07785/) (2021)
- [Bilingual Rhetorical Structure Parsing with Large Parallel Annotations](https://mlsanity.com/paper/2409.14969/) (2024)
- [Generating Related Work](https://mlsanity.com/paper/2104.08668/) (2021)
- [Less is More: A Lightweight and Robust Neural Architecture for Discourse Parsing](https://mlsanity.com/paper/2210.09537/) (2022)
- [RST-style Discourse Parsing Guided by Document-level Content Structures](https://mlsanity.com/paper/2309.04141/) (2023)

## Abstract (arXiv)

Rhetorical structure trees have been shown to be useful for several document-level tasks including summarization and document classification. Previous approaches to RST parsing have used discriminative models; however, these are less sample efficient than generative models, and RST parsing datasets are typically small. In this paper, we present the first generative model for RST parsing. Our model is a document-level RNN grammar (RNNG) with a bottom-up traversal order. We show that, for our parser's traversal order, previous beam search algorithms for RNNGs have a left-branching bias which is ill-suited for RST parsing. We develop a novel beam search algorithm that keeps track of both structure- and word-generating actions without exhibiting this branching bias and results in absolute improvements of 6.8 and 2.9 on unlabelled and labelled F1 over previous algorithms. Overall, our generative model outperforms a discriminative model with the same features by 2.6 F1 points and achieves performance comparable to the state-of-the-art, outperforming all published parsers from a recent replication study that do not use additional training data.