EUVIMEDEUVIMED entdecken
Crossref

Discourse-Aware Language Representation

Zhuosheng Zhang, Siru Ouyang, Hai Zhao

IEEE Transactions on Pattern Analysis and Machine Intelligence · 2026-02 · 48 (2) · 1888-1903

DOI: 10.1109/tpami.2025.3621229 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Pattern Analysis and Machine Intelligence, 2026-02. Verfasst von Zhuosheng Zhang und 2 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 117 Literaturangaben. Gefördert von: Chinese National Foundation of Social Sciences, Challenge and Governance of Smart Media on News Authenticity, National Natural Science Foundation of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Recent Transformer-based language representation techniques have commonly adopted a straightforward approach to modeling textual context as a linear sequence of successive tokens. However, this sequential modeling strategy falls short in actively exploring intermediate structures present in natural languages and does not account for the rich interactive relationships between sentences. To overcome these limitations, we propose a discourse-aware framework that bridges the gap between sequential contextualization and the interactive nature of conversational reading comprehension. Concretely, we first divide the context into elementary discourse units (EDUs), ensuring that each unit contains precisely one condition. Then, we systematically explore three instantiations for modeling discourse features: sequential EDU encoding, discourse-aware masking, and discourse graph network. These techniques allow us to capture the nuanced interactions within the discourse. To assess the efficacy of our methodologies, we perform experiments on three conversational reading comprehension tasks: multi-turn response selection, conversational question answering, and conversational machine reading. Experimental results demonstrate the superiority of our proposed approach. Moreover, analysis reveals that the discourse-aware approach enables the model to effectively capture intricate relationships within the context and fosters reasoning interpretability. Additionally, our method exhibits efficacy across various backbone PLMs and diverse domains.

Förderung

  • Chinese National Foundation of Social Sciences
  • Challenge and Governance of Smart Media on News Authenticity
  • National Natural Science Foundation of China
  • Natural Science Foundation of Shanghai

Zitieren

APA

Zhang, Z., Ouyang, S., & Zhao, H. (2026). Discourse-Aware Language Representation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 48(2), 1888-1903. https://doi.org/10.1109/tpami.2025.3621229

Vancouver

Zhang Z, Ouyang S, Zhao H. Discourse-Aware Language Representation. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2026;48(2):1888-1903. doi:10.1109/tpami.2025.3621229

BibTeX
@article{Zhang2026,
  author = {Zhang, Zhuosheng and Ouyang, Siru and Zhao, Hai},
  title = {Discourse-Aware Language Representation},
  journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},
  year = {2026},
  volume = {48},
  number = {2},
  pages = {1888--1903},
  doi = {10.1109/tpami.2025.3621229}
}
Kostenlos starten →