@inproceedings{braunger-etal-2016-comparative,
title = "A Comparative Analysis of Crowdsourced Natural Language Corpora for Spoken Dialog Systems",
author = {Braunger, Patricia and
Hofmann, Hansj{\"o}rg and
Werner, Steffen and
Schmidt, Maria},
booktitle = "Proceedings of the Tenth International Conference on Language Resources and Evaluation ({LREC}'16)",
month = may,
year = "2016",
address = "Portoro{\v{z}}, Slovenia",
publisher = "European Language Resources Association (ELRA)",
url = "https://aclanthology.org/L16-1119",
pages = "750--755",
abstract = "Recent spoken dialog systems have been able to recognize freely spoken user input in restricted domains thanks to statistical methods in the automatic speech recognition. These methods require a high number of natural language utterances to train the speech recognition engine and to assess the quality of the system. Since human speech offers many variants associated with a single intent, a high number of user utterances have to be elicited. Developers are therefore turning to crowdsourcing to collect this data. This paper compares three different methods to elicit multiple utterances for given semantics via crowd sourcing, namely with pictures, with text and with semantic entities. Specifically, we compare the methods with regard to the number of valid data and linguistic variance, whereby a quantitative and qualitative approach is proposed. In our study, the method with text led to a high variance in the utterances and a relatively low rate of invalid data.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="braunger-etal-2016-comparative">
<titleInfo>
<title>A Comparative Analysis of Crowdsourced Natural Language Corpora for Spoken Dialog Systems</title>
</titleInfo>
<name type="personal">
<namePart type="given">Patricia</namePart>
<namePart type="family">Braunger</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hansjörg</namePart>
<namePart type="family">Hofmann</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Steffen</namePart>
<namePart type="family">Werner</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Maria</namePart>
<namePart type="family">Schmidt</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2016-may</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Tenth International Conference on Language Resources and Evaluation (LREC’16)</title>
</titleInfo>
<originInfo>
<publisher>European Language Resources Association (ELRA)</publisher>
<place>
<placeTerm type="text">Portorož, Slovenia</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Recent spoken dialog systems have been able to recognize freely spoken user input in restricted domains thanks to statistical methods in the automatic speech recognition. These methods require a high number of natural language utterances to train the speech recognition engine and to assess the quality of the system. Since human speech offers many variants associated with a single intent, a high number of user utterances have to be elicited. Developers are therefore turning to crowdsourcing to collect this data. This paper compares three different methods to elicit multiple utterances for given semantics via crowd sourcing, namely with pictures, with text and with semantic entities. Specifically, we compare the methods with regard to the number of valid data and linguistic variance, whereby a quantitative and qualitative approach is proposed. In our study, the method with text led to a high variance in the utterances and a relatively low rate of invalid data.</abstract>
<identifier type="citekey">braunger-etal-2016-comparative</identifier>
<location>
<url>https://aclanthology.org/L16-1119</url>
</location>
<part>
<date>2016-may</date>
<extent unit="page">
<start>750</start>
<end>755</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T A Comparative Analysis of Crowdsourced Natural Language Corpora for Spoken Dialog Systems
%A Braunger, Patricia
%A Hofmann, Hansjörg
%A Werner, Steffen
%A Schmidt, Maria
%S Proceedings of the Tenth International Conference on Language Resources and Evaluation (LREC’16)
%D 2016
%8 may
%I European Language Resources Association (ELRA)
%C Portorož, Slovenia
%F braunger-etal-2016-comparative
%X Recent spoken dialog systems have been able to recognize freely spoken user input in restricted domains thanks to statistical methods in the automatic speech recognition. These methods require a high number of natural language utterances to train the speech recognition engine and to assess the quality of the system. Since human speech offers many variants associated with a single intent, a high number of user utterances have to be elicited. Developers are therefore turning to crowdsourcing to collect this data. This paper compares three different methods to elicit multiple utterances for given semantics via crowd sourcing, namely with pictures, with text and with semantic entities. Specifically, we compare the methods with regard to the number of valid data and linguistic variance, whereby a quantitative and qualitative approach is proposed. In our study, the method with text led to a high variance in the utterances and a relatively low rate of invalid data.
%U https://aclanthology.org/L16-1119
%P 750-755
Markdown (Informal)
[A Comparative Analysis of Crowdsourced Natural Language Corpora for Spoken Dialog Systems](https://aclanthology.org/L16-1119) (Braunger et al., LREC 2016)
ACL