@inproceedings{jiang-de-rijke-2018-sequence,
title = "Why are Sequence-to-Sequence Models So Dull? Understanding the Low-Diversity Problem of Chatbots",
author = "Jiang, Shaojie and
de Rijke, Maarten",
editor = "Chuklin, Aleksandr and
Dalton, Jeff and
Kiseleva, Julia and
Borisov, Alexey and
Burtsev, Mikhail",
booktitle = "Proceedings of the 2018 {EMNLP} Workshop {SCAI}: The 2nd International Workshop on Search-Oriented Conversational {AI}",
month = oct,
year = "2018",
address = "Brussels, Belgium",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/W18-5712",
doi = "10.18653/v1/W18-5712",
pages = "81--86",
abstract = "Diversity is a long-studied topic in information retrieval that usually refers to the requirement that retrieved results should be non-repetitive and cover different aspects. In a conversational setting, an additional dimension of diversity matters: an engaging response generation system should be able to output responses that are diverse and interesting. Sequence-to-sequence (Seq2Seq) models have been shown to be very effective for response generation. However, dialogue responses generated by Seq2Seq models tend to have low diversity. In this paper, we review known sources and existing approaches to this low-diversity problem. We also identify a source of low diversity that has been little studied so far, namely model over-confidence. We sketch several directions for tackling model over-confidence and, hence, the low-diversity problem, including confidence penalties and label smoothing.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="jiang-de-rijke-2018-sequence">
<titleInfo>
<title>Why are Sequence-to-Sequence Models So Dull? Understanding the Low-Diversity Problem of Chatbots</title>
</titleInfo>
<name type="personal">
<namePart type="given">Shaojie</namePart>
<namePart type="family">Jiang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Maarten</namePart>
<namePart type="family">de Rijke</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2018-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2018 EMNLP Workshop SCAI: The 2nd International Workshop on Search-Oriented Conversational AI</title>
</titleInfo>
<name type="personal">
<namePart type="given">Aleksandr</namePart>
<namePart type="family">Chuklin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jeff</namePart>
<namePart type="family">Dalton</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Julia</namePart>
<namePart type="family">Kiseleva</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alexey</namePart>
<namePart type="family">Borisov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mikhail</namePart>
<namePart type="family">Burtsev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Brussels, Belgium</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Diversity is a long-studied topic in information retrieval that usually refers to the requirement that retrieved results should be non-repetitive and cover different aspects. In a conversational setting, an additional dimension of diversity matters: an engaging response generation system should be able to output responses that are diverse and interesting. Sequence-to-sequence (Seq2Seq) models have been shown to be very effective for response generation. However, dialogue responses generated by Seq2Seq models tend to have low diversity. In this paper, we review known sources and existing approaches to this low-diversity problem. We also identify a source of low diversity that has been little studied so far, namely model over-confidence. We sketch several directions for tackling model over-confidence and, hence, the low-diversity problem, including confidence penalties and label smoothing.</abstract>
<identifier type="citekey">jiang-de-rijke-2018-sequence</identifier>
<identifier type="doi">10.18653/v1/W18-5712</identifier>
<location>
<url>https://aclanthology.org/W18-5712</url>
</location>
<part>
<date>2018-10</date>
<extent unit="page">
<start>81</start>
<end>86</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Why are Sequence-to-Sequence Models So Dull? Understanding the Low-Diversity Problem of Chatbots
%A Jiang, Shaojie
%A de Rijke, Maarten
%Y Chuklin, Aleksandr
%Y Dalton, Jeff
%Y Kiseleva, Julia
%Y Borisov, Alexey
%Y Burtsev, Mikhail
%S Proceedings of the 2018 EMNLP Workshop SCAI: The 2nd International Workshop on Search-Oriented Conversational AI
%D 2018
%8 October
%I Association for Computational Linguistics
%C Brussels, Belgium
%F jiang-de-rijke-2018-sequence
%X Diversity is a long-studied topic in information retrieval that usually refers to the requirement that retrieved results should be non-repetitive and cover different aspects. In a conversational setting, an additional dimension of diversity matters: an engaging response generation system should be able to output responses that are diverse and interesting. Sequence-to-sequence (Seq2Seq) models have been shown to be very effective for response generation. However, dialogue responses generated by Seq2Seq models tend to have low diversity. In this paper, we review known sources and existing approaches to this low-diversity problem. We also identify a source of low diversity that has been little studied so far, namely model over-confidence. We sketch several directions for tackling model over-confidence and, hence, the low-diversity problem, including confidence penalties and label smoothing.
%R 10.18653/v1/W18-5712
%U https://aclanthology.org/W18-5712
%U https://doi.org/10.18653/v1/W18-5712
%P 81-86
Markdown (Informal)
[Why are Sequence-to-Sequence Models So Dull? Understanding the Low-Diversity Problem of Chatbots](https://aclanthology.org/W18-5712) (Jiang & de Rijke, EMNLP 2018)
ACL