@inproceedings{czeresnia-etinger-black-2019-formality,
title = "Formality Style Transfer for Noisy, User-generated Conversations: Extracting Labeled, Parallel Data from Unlabeled Corpora",
author = "Czeresnia Etinger, Isak and
Black, Alan W",
editor = "Xu, Wei and
Ritter, Alan and
Baldwin, Tim and
Rahimi, Afshin",
booktitle = "Proceedings of the 5th Workshop on Noisy User-generated Text (W-NUT 2019)",
month = nov,
year = "2019",
address = "Hong Kong, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/D19-5502",
doi = "10.18653/v1/D19-5502",
pages = "11--16",
abstract = "Typical datasets used for style transfer in NLP contain aligned pairs of two opposite extremes of a style. As each existing dataset is sourced from a specific domain and context, most use cases will have a sizable mismatch from the vocabulary and sentence structures of any dataset available. This reduces the performance of the style transfer, and is particularly significant for noisy, user-generated text. To solve this problem, we show a technique to derive a dataset of aligned pairs (style-agnostic vs stylistic sentences) from an unlabeled corpus by using an auxiliary dataset, allowing for in-domain training. We test the technique with the Yahoo Formality Dataset and 6 novel datasets we produced, which consist of scripts from 5 popular TV-shows (Friends, Futurama, Seinfeld, Southpark, Stargate SG-1) and the Slate Star Codex online forum. We gather 1080 human evaluations, which show that our method produces a sizable change in formality while maintaining fluency and context; and that it considerably outperforms OpenNMT{'}s Seq2Seq model directly trained on the Yahoo Formality Dataset. Additionally, we publish the full pipeline code and our novel datasets.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="czeresnia-etinger-black-2019-formality">
<titleInfo>
<title>Formality Style Transfer for Noisy, User-generated Conversations: Extracting Labeled, Parallel Data from Unlabeled Corpora</title>
</titleInfo>
<name type="personal">
<namePart type="given">Isak</namePart>
<namePart type="family">Czeresnia Etinger</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alan</namePart>
<namePart type="given">W</namePart>
<namePart type="family">Black</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2019-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 5th Workshop on Noisy User-generated Text (W-NUT 2019)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wei</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alan</namePart>
<namePart type="family">Ritter</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tim</namePart>
<namePart type="family">Baldwin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Afshin</namePart>
<namePart type="family">Rahimi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Hong Kong, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Typical datasets used for style transfer in NLP contain aligned pairs of two opposite extremes of a style. As each existing dataset is sourced from a specific domain and context, most use cases will have a sizable mismatch from the vocabulary and sentence structures of any dataset available. This reduces the performance of the style transfer, and is particularly significant for noisy, user-generated text. To solve this problem, we show a technique to derive a dataset of aligned pairs (style-agnostic vs stylistic sentences) from an unlabeled corpus by using an auxiliary dataset, allowing for in-domain training. We test the technique with the Yahoo Formality Dataset and 6 novel datasets we produced, which consist of scripts from 5 popular TV-shows (Friends, Futurama, Seinfeld, Southpark, Stargate SG-1) and the Slate Star Codex online forum. We gather 1080 human evaluations, which show that our method produces a sizable change in formality while maintaining fluency and context; and that it considerably outperforms OpenNMT’s Seq2Seq model directly trained on the Yahoo Formality Dataset. Additionally, we publish the full pipeline code and our novel datasets.</abstract>
<identifier type="citekey">czeresnia-etinger-black-2019-formality</identifier>
<identifier type="doi">10.18653/v1/D19-5502</identifier>
<location>
<url>https://aclanthology.org/D19-5502</url>
</location>
<part>
<date>2019-11</date>
<extent unit="page">
<start>11</start>
<end>16</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Formality Style Transfer for Noisy, User-generated Conversations: Extracting Labeled, Parallel Data from Unlabeled Corpora
%A Czeresnia Etinger, Isak
%A Black, Alan W.
%Y Xu, Wei
%Y Ritter, Alan
%Y Baldwin, Tim
%Y Rahimi, Afshin
%S Proceedings of the 5th Workshop on Noisy User-generated Text (W-NUT 2019)
%D 2019
%8 November
%I Association for Computational Linguistics
%C Hong Kong, China
%F czeresnia-etinger-black-2019-formality
%X Typical datasets used for style transfer in NLP contain aligned pairs of two opposite extremes of a style. As each existing dataset is sourced from a specific domain and context, most use cases will have a sizable mismatch from the vocabulary and sentence structures of any dataset available. This reduces the performance of the style transfer, and is particularly significant for noisy, user-generated text. To solve this problem, we show a technique to derive a dataset of aligned pairs (style-agnostic vs stylistic sentences) from an unlabeled corpus by using an auxiliary dataset, allowing for in-domain training. We test the technique with the Yahoo Formality Dataset and 6 novel datasets we produced, which consist of scripts from 5 popular TV-shows (Friends, Futurama, Seinfeld, Southpark, Stargate SG-1) and the Slate Star Codex online forum. We gather 1080 human evaluations, which show that our method produces a sizable change in formality while maintaining fluency and context; and that it considerably outperforms OpenNMT’s Seq2Seq model directly trained on the Yahoo Formality Dataset. Additionally, we publish the full pipeline code and our novel datasets.
%R 10.18653/v1/D19-5502
%U https://aclanthology.org/D19-5502
%U https://doi.org/10.18653/v1/D19-5502
%P 11-16
Markdown (Informal)
[Formality Style Transfer for Noisy, User-generated Conversations: Extracting Labeled, Parallel Data from Unlabeled Corpora](https://aclanthology.org/D19-5502) (Czeresnia Etinger & Black, WNUT 2019)
ACL