@inproceedings{shimorina-gardent-2018-handling,
title = "Handling Rare Items in Data-to-Text Generation",
author = "Shimorina, Anastasia and
Gardent, Claire",
editor = "Krahmer, Emiel and
Gatt, Albert and
Goudbeek, Martijn",
booktitle = "Proceedings of the 11th International Conference on Natural Language Generation",
month = nov,
year = "2018",
address = "Tilburg University, The Netherlands",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/W18-6543",
doi = "10.18653/v1/W18-6543",
pages = "360--370",
abstract = "Neural approaches to data-to-text generation generally handle rare input items using either delexicalisation or a copy mechanism. We investigate the relative impact of these two methods on two datasets (E2E and WebNLG) and using two evaluation settings. We show (i) that rare items strongly impact performance; (ii) that combining delexicalisation and copying yields the strongest improvement; (iii) that copying underperforms for rare and unseen items and (iv) that the impact of these two mechanisms greatly varies depending on how the dataset is constructed and on how it is split into train, dev and test.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="shimorina-gardent-2018-handling">
<titleInfo>
<title>Handling Rare Items in Data-to-Text Generation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Anastasia</namePart>
<namePart type="family">Shimorina</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Claire</namePart>
<namePart type="family">Gardent</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2018-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 11th International Conference on Natural Language Generation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Emiel</namePart>
<namePart type="family">Krahmer</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Albert</namePart>
<namePart type="family">Gatt</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Martijn</namePart>
<namePart type="family">Goudbeek</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Tilburg University, The Netherlands</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Neural approaches to data-to-text generation generally handle rare input items using either delexicalisation or a copy mechanism. We investigate the relative impact of these two methods on two datasets (E2E and WebNLG) and using two evaluation settings. We show (i) that rare items strongly impact performance; (ii) that combining delexicalisation and copying yields the strongest improvement; (iii) that copying underperforms for rare and unseen items and (iv) that the impact of these two mechanisms greatly varies depending on how the dataset is constructed and on how it is split into train, dev and test.</abstract>
<identifier type="citekey">shimorina-gardent-2018-handling</identifier>
<identifier type="doi">10.18653/v1/W18-6543</identifier>
<location>
<url>https://aclanthology.org/W18-6543</url>
</location>
<part>
<date>2018-11</date>
<extent unit="page">
<start>360</start>
<end>370</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Handling Rare Items in Data-to-Text Generation
%A Shimorina, Anastasia
%A Gardent, Claire
%Y Krahmer, Emiel
%Y Gatt, Albert
%Y Goudbeek, Martijn
%S Proceedings of the 11th International Conference on Natural Language Generation
%D 2018
%8 November
%I Association for Computational Linguistics
%C Tilburg University, The Netherlands
%F shimorina-gardent-2018-handling
%X Neural approaches to data-to-text generation generally handle rare input items using either delexicalisation or a copy mechanism. We investigate the relative impact of these two methods on two datasets (E2E and WebNLG) and using two evaluation settings. We show (i) that rare items strongly impact performance; (ii) that combining delexicalisation and copying yields the strongest improvement; (iii) that copying underperforms for rare and unseen items and (iv) that the impact of these two mechanisms greatly varies depending on how the dataset is constructed and on how it is split into train, dev and test.
%R 10.18653/v1/W18-6543
%U https://aclanthology.org/W18-6543
%U https://doi.org/10.18653/v1/W18-6543
%P 360-370
Markdown (Informal)
[Handling Rare Items in Data-to-Text Generation](https://aclanthology.org/W18-6543) (Shimorina & Gardent, INLG 2018)
ACL
- Anastasia Shimorina and Claire Gardent. 2018. Handling Rare Items in Data-to-Text Generation. In Proceedings of the 11th International Conference on Natural Language Generation, pages 360–370, Tilburg University, The Netherlands. Association for Computational Linguistics.