State-of-the-art generalisation research in NLP: a taxonomy and review

D Hupkes, M Giulianelli, V Dankers, M Artetxe… - arXiv preprint arXiv …, 2022 - arxiv.org
The ability to generalise well is one of the primary desiderata of natural language
processing (NLP). Yet, what'good generalisation'entails and how it should be evaluated is …

DisCut and DiscReT: MELODI at DISRPT 2023

E Metheniti, C Braud, P Muller, L Rivière - 3rd Shared Task on …, 2023 - hal.science
This paper presents the results obtained by the MELODI team for the three tasks proposed
within the DISRPT 2023 shared task on discourse: segmentation, connective identification …

Why can't discourse parsing generalize? A thorough investigation of the impact of data diversity

YJ Liu, A Zeldes - arXiv preprint arXiv:2302.06488, 2023 - arxiv.org
Recent advances in discourse parsing performance create the impression that, as in other
NLP tasks, performance for high-resource languages such as English is finally becoming …

Rst discourse parsing as text-to-text generation

X Hu, X Wan - IEEE/ACM Transactions on Audio, Speech, and …, 2023 - ieeexplore.ieee.org
Previous studies have made great advances in RST discourse parsing through specific
neural frameworks or features, but they usually split the parsing process into two subtasks …

Incorporating distributions of discourse structure for long document abstractive summarization

D Pu, Y Wang, V Demberg - arXiv preprint arXiv:2305.16784, 2023 - arxiv.org
For text summarization, the role of discourse structure is pivotal in discerning the core
content of a text. Regrettably, prior studies on incorporating Rhetorical Structure Theory …

GCDT: A Chinese RST treebank for multigenre and multilingual discourse parsing

S Peng, YJ Liu, A Zeldes - arXiv preprint arXiv:2210.10449, 2022 - arxiv.org
A lack of large-scale human-annotated data has hampered the hierarchical discourse
parsing of Chinese. In this paper, we present GCDT, the largest hierarchical discourse …

[HTML][HTML] Generating EDU extracts for plan-guided summary re-ranking

G Adams, AR Fabbri, F Ladhak… - Proceedings of the …, 2023 - ncbi.nlm.nih.gov
Two-step approaches, in which summary candidates are generated-then-reranked to return
a single summary, can improve ROUGE scores over the standard single-step approach. Yet …

What's Hard in English RST Parsing? Predictive Models for Error Analysis

YJ Liu, T Aoyama, A Zeldes - arXiv preprint arXiv:2309.04940, 2023 - arxiv.org
Despite recent advances in Natural Language Processing (NLP), hierarchical discourse
parsing in the framework of Rhetorical Structure Theory remains challenging, and our …

RST-LoRA: A Discourse-Aware Low-Rank Adaptation for Long Document Abstractive Summarization

D Pu, V Demberg - arXiv preprint arXiv:2405.00657, 2024 - arxiv.org
For long document summarization, discourse structure is important to discern the key
content of the text and the differences in importance level between sentences. Unfortunately …

End-to-end argument mining over varying rhetorical structures

E Chistova - arXiv preprint arXiv:2401.11218, 2024 - arxiv.org
Rhetorical Structure Theory implies no single discourse interpretation of a text, and the
limitations of RST parsers further exacerbate inconsistent parsing of similar structures …