Forskning ved Københavns Universitet - Københavns Universitet

Forside

X-WikiRE: A Large, Multilingual Resource for Relation Extraction as Machine Comprehension

Publikation: Bidrag til bog/antologi/rapportKonferencebidrag i proceedingsForskningfagfællebedømt

Dokumenter

  • OA-X-WikiRE

    Forlagets udgivne version, 1,87 MB, PDF-dokument

Although the vast majority of knowledge bases (KBs) are heavily biased towards English, Wikipedias do cover very different topics in different languages. Exploiting this, we introduce a new multilingual dataset (X-WikiRE), framing relation extraction as a multilingual machine reading problem. We show that by leveraging this resource it is possible to robustly transfer models cross-lingually and that multilingual support significantly improves (zero-shot) relation extraction, enabling the population of low-resourced KBs from their well-populated counterparts.
OriginalsprogEngelsk
TitelProceedings of the 2nd Workshop on Deep Learning Approaches for Low-Resource NLP (DeepLo 2019)
ForlagAssociation for Computational Linguistics
Publikationsdato2019
Sider265-274
DOI
StatusUdgivet - 2019
Begivenhed2nd Workshop on Deep Learning Approaches for Low-Resource NLP (DeepLo 2019) - Hong Kong, China
Varighed: 1 nov. 20191 nov. 2019

Konference

Konference2nd Workshop on Deep Learning Approaches for Low-Resource NLP (DeepLo 2019)
LokationHong Kong, China
Periode01/11/201901/11/2019

Antal downloads er baseret på statistik fra Google Scholar og www.ku.dk


Ingen data tilgængelig

ID: 239861905