<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="hsueh-ma-2020-semantic">
    <titleInfo>
        <title>Semantic Guidance of Dialogue Generation with Reinforcement Learning</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Cheng-Hsun</namePart>
        <namePart type="family">Hsueh</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Wei-Yun</namePart>
        <namePart type="family">Ma</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2020-07</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 21th Annual Meeting of the Special Interest Group on Discourse and Dialogue</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Olivier</namePart>
            <namePart type="family">Pietquin</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Smaranda</namePart>
            <namePart type="family">Muresan</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Vivian</namePart>
            <namePart type="family">Chen</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Casey</namePart>
            <namePart type="family">Kennington</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">David</namePart>
            <namePart type="family">Vandyke</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Nina</namePart>
            <namePart type="family">Dethlefs</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Koji</namePart>
            <namePart type="family">Inoue</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Erik</namePart>
            <namePart type="family">Ekstedt</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Stefan</namePart>
            <namePart type="family">Ultes</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">1st virtual meeting</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
    </relatedItem>
    <abstract>Neural encoder-decoder models have shown promising performance for human-computer dialogue systems over the past few years. However, due to the maximum-likelihood objective for the decoder, the generated responses are often universal and safe to the point that they lack meaningful information and are no longer relevant to the post. To address this, in this paper, we propose semantic guidance using reinforcement learning to ensure that the generated responses indeed include the given or predicted semantics and that these semantics do not appear repeatedly in the response. Synsets, which comprise sets of manually defined synonyms, are used as the form of assigned semantics. For a given/assigned/predicted synset, only one of its synonyms should appear in the generated response; this constitutes a simple but effective semantic-control mechanism. We conduct both quantitative and qualitative evaluations, which show that the generated responses are not only higher-quality but also reflect the assigned semantic controls.</abstract>
    <identifier type="citekey">hsueh-ma-2020-semantic</identifier>
    <identifier type="doi">10.18653/v1/2020.sigdial-1.1</identifier>
    <location>
        <url>https://aclanthology.org/2020.sigdial-1.1</url>
    </location>
    <part>
        <date>2020-07</date>
        <extent unit="page">
            <start>1</start>
            <end>9</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Semantic Guidance of Dialogue Generation with Reinforcement Learning
%A Hsueh, Cheng-Hsun
%A Ma, Wei-Yun
%Y Pietquin, Olivier
%Y Muresan, Smaranda
%Y Chen, Vivian
%Y Kennington, Casey
%Y Vandyke, David
%Y Dethlefs, Nina
%Y Inoue, Koji
%Y Ekstedt, Erik
%Y Ultes, Stefan
%S Proceedings of the 21th Annual Meeting of the Special Interest Group on Discourse and Dialogue
%D 2020
%8 July
%I Association for Computational Linguistics
%C 1st virtual meeting
%F hsueh-ma-2020-semantic
%X Neural encoder-decoder models have shown promising performance for human-computer dialogue systems over the past few years. However, due to the maximum-likelihood objective for the decoder, the generated responses are often universal and safe to the point that they lack meaningful information and are no longer relevant to the post. To address this, in this paper, we propose semantic guidance using reinforcement learning to ensure that the generated responses indeed include the given or predicted semantics and that these semantics do not appear repeatedly in the response. Synsets, which comprise sets of manually defined synonyms, are used as the form of assigned semantics. For a given/assigned/predicted synset, only one of its synonyms should appear in the generated response; this constitutes a simple but effective semantic-control mechanism. We conduct both quantitative and qualitative evaluations, which show that the generated responses are not only higher-quality but also reflect the assigned semantic controls.
%R 10.18653/v1/2020.sigdial-1.1
%U https://aclanthology.org/2020.sigdial-1.1
%U https://doi.org/10.18653/v1/2020.sigdial-1.1
%P 1-9

Markdown (Informal)

[Semantic Guidance of Dialogue Generation with Reinforcement Learning](https://aclanthology.org/2020.sigdial-1.1) (Hsueh & Ma, SIGDIAL 2020)

• Semantic Guidance of Dialogue Generation with Reinforcement Learning (Hsueh & Ma, SIGDIAL 2020)

ACL

• Cheng-Hsun Hsueh and Wei-Yun Ma. 2020. Semantic Guidance of Dialogue Generation with Reinforcement Learning. In Proceedings of the 21th Annual Meeting of the Special Interest Group on Discourse and Dialogue, pages 1–9, 1st virtual meeting. Association for Computational Linguistics.