@inproceedings{hartmann-etal-2017-assessing,
title = "Assessing {SRL} Frameworks with Automatic Training Data Expansion",
author = "Hartmann, Silvana and
M{\'u}jdricza-Maydt, {\'E}va and
Kuznetsov, Ilia and
Gurevych, Iryna and
Frank, Anette",
editor = "Schneider, Nathan and
Xue, Nianwen",
booktitle = "Proceedings of the 11th Linguistic Annotation Workshop",
month = apr,
year = "2017",
address = "Valencia, Spain",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/W17-0814",
doi = "10.18653/v1/W17-0814",
pages = "115--121",
abstract = "We present the first experiment-based study that explicitly contrasts the three major semantic role labeling frameworks. As a prerequisite, we create a dataset labeled with parallel FrameNet-, PropBank-, and VerbNet-style labels for German. We train a state-of-the-art SRL tool for German for the different annotation styles and provide a comparative analysis across frameworks. We further explore the behavior of the frameworks with automatic training data generation. VerbNet provides larger semantic expressivity than PropBank, and we find that its generalization capacity approaches PropBank in SRL training, but it benefits less from training data expansion than the sparse-data affected FrameNet.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="hartmann-etal-2017-assessing">
<titleInfo>
<title>Assessing SRL Frameworks with Automatic Training Data Expansion</title>
</titleInfo>
<name type="personal">
<namePart type="given">Silvana</namePart>
<namePart type="family">Hartmann</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Éva</namePart>
<namePart type="family">Mújdricza-Maydt</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ilia</namePart>
<namePart type="family">Kuznetsov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Iryna</namePart>
<namePart type="family">Gurevych</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Anette</namePart>
<namePart type="family">Frank</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2017-04</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 11th Linguistic Annotation Workshop</title>
</titleInfo>
<name type="personal">
<namePart type="given">Nathan</namePart>
<namePart type="family">Schneider</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nianwen</namePart>
<namePart type="family">Xue</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Valencia, Spain</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>We present the first experiment-based study that explicitly contrasts the three major semantic role labeling frameworks. As a prerequisite, we create a dataset labeled with parallel FrameNet-, PropBank-, and VerbNet-style labels for German. We train a state-of-the-art SRL tool for German for the different annotation styles and provide a comparative analysis across frameworks. We further explore the behavior of the frameworks with automatic training data generation. VerbNet provides larger semantic expressivity than PropBank, and we find that its generalization capacity approaches PropBank in SRL training, but it benefits less from training data expansion than the sparse-data affected FrameNet.</abstract>
<identifier type="citekey">hartmann-etal-2017-assessing</identifier>
<identifier type="doi">10.18653/v1/W17-0814</identifier>
<location>
<url>https://aclanthology.org/W17-0814</url>
</location>
<part>
<date>2017-04</date>
<extent unit="page">
<start>115</start>
<end>121</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Assessing SRL Frameworks with Automatic Training Data Expansion
%A Hartmann, Silvana
%A Mújdricza-Maydt, Éva
%A Kuznetsov, Ilia
%A Gurevych, Iryna
%A Frank, Anette
%Y Schneider, Nathan
%Y Xue, Nianwen
%S Proceedings of the 11th Linguistic Annotation Workshop
%D 2017
%8 April
%I Association for Computational Linguistics
%C Valencia, Spain
%F hartmann-etal-2017-assessing
%X We present the first experiment-based study that explicitly contrasts the three major semantic role labeling frameworks. As a prerequisite, we create a dataset labeled with parallel FrameNet-, PropBank-, and VerbNet-style labels for German. We train a state-of-the-art SRL tool for German for the different annotation styles and provide a comparative analysis across frameworks. We further explore the behavior of the frameworks with automatic training data generation. VerbNet provides larger semantic expressivity than PropBank, and we find that its generalization capacity approaches PropBank in SRL training, but it benefits less from training data expansion than the sparse-data affected FrameNet.
%R 10.18653/v1/W17-0814
%U https://aclanthology.org/W17-0814
%U https://doi.org/10.18653/v1/W17-0814
%P 115-121
Markdown (Informal)
[Assessing SRL Frameworks with Automatic Training Data Expansion](https://aclanthology.org/W17-0814) (Hartmann et al., LAW 2017)
ACL