Stage-oe-small.jpg

Inproceedings3943

Aus Aifbportal
Version vom 4. Juli 2022, 08:19 Uhr von Jg2894 (Diskussion | Beiträge)
(Unterschied) ← Nächstältere Version | Aktuelle Version (Unterschied) | Nächstjüngere Version → (Unterschied)
Wechseln zu:Navigation, Suche


Few-Shot Document-Level Relation Extraction


Few-Shot Document-Level Relation Extraction



Published: 2022

Buchtitel: Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies
Verlag: Association for Computational Linguistics

Referierte Veröffentlichung

BibTeX

Kurzfassung
We present FREDo, a few-shot document-level relation extraction (FSDLRE) benchmark. As opposed to existing benchmarks which are built on sentence-level relation extraction corpora, we argue that document-level corpora provide more realism, particularly regarding none-of-the-above (NOTA) distributions. Therefore, we propose a set of FSDLRE tasks and construct a benchmark based on two existing supervised learning data sets, DocRED and sciERC. We adapt the state-of-the-art sentence-level method MNAV to the document-level and develop it further for improved domain adaptation. We find FSDLRE to be a challenging setting with interesting new characteristics such as the ability to sample NOTA instances from the support set. The data, code, and trained models are available online (https://github.com/nicpopovic/FREDo).

Download: Media:FSDLRE.pdf



Forschungsgruppe

Web Science


Forschungsgebiet

Text Mining, Informationsextraktion, Natürliche Sprachverarbeitung, Deep Learning