@inproceedings{kim-etal-2024-revealing,
title = "Revealing User Familiarity Bias in Task-Oriented Dialogue via Interactive Evaluation",
author = "Kim, Takyoung and
Shin, Jamin and
Kim, Young-Ho and
Bae, Sanghwan and
Kim, Sungdong",
editor = "Nouri, Elnaz and
Rastogi, Abhinav and
Spithourakis, Georgios and
Liu, Bing and
Chen, Yun-Nung and
Li, Yu and
Albalak, Alon and
Wakaki, Hiromi and
Papangelis, Alexandros",
booktitle = "Proceedings of the 6th Workshop on NLP for Conversational AI (NLP4ConvAI 2024)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.nlp4convai-1.3",
pages = "37--55",
abstract = "Most task-oriented dialogue (TOD) benchmarks assume users that know exactly how to use the system by constraining the user behaviors within the system{'}s capabilities via strict user goals, namely {``}user familiarity{''} bias. This data bias deepens when it combines with data-driven TOD systems, as it is impossible to fathom the effect of it with existing static evaluations. Hence, we conduct an interactive user study to unveil how vulnerable TOD systems are against realistic scenarios. In particular, we compare users with 1) detailed goal instructions that conform to the system boundaries (closed-goal) and 2) vague goal instructions that are often unsupported but realistic (open-goal). Our study reveals that conversations in open-goal settings lead to catastrophic failures of the system, in which 92{\%} of the dialogues had significant issues. Moreover, we conduct a thorough analysis to identify distinctive features between the two settings through error annotation. From this, we discover a novel {``}pretending{''} behavior, in which the system pretends to handle the user requests even though they are beyond the system{'}s capabilities. We discuss its characteristics and toxicity while showing recent large language models can also suffer from this behavior.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="kim-etal-2024-revealing">
<titleInfo>
<title>Revealing User Familiarity Bias in Task-Oriented Dialogue via Interactive Evaluation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Takyoung</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jamin</namePart>
<namePart type="family">Shin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Young-Ho</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sanghwan</namePart>
<namePart type="family">Bae</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sungdong</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-08</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 6th Workshop on NLP for Conversational AI (NLP4ConvAI 2024)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Elnaz</namePart>
<namePart type="family">Nouri</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Abhinav</namePart>
<namePart type="family">Rastogi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Georgios</namePart>
<namePart type="family">Spithourakis</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bing</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yun-Nung</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yu</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alon</namePart>
<namePart type="family">Albalak</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hiromi</namePart>
<namePart type="family">Wakaki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alexandros</namePart>
<namePart type="family">Papangelis</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Bangkok, Thailand</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Most task-oriented dialogue (TOD) benchmarks assume users that know exactly how to use the system by constraining the user behaviors within the system’s capabilities via strict user goals, namely “user familiarity” bias. This data bias deepens when it combines with data-driven TOD systems, as it is impossible to fathom the effect of it with existing static evaluations. Hence, we conduct an interactive user study to unveil how vulnerable TOD systems are against realistic scenarios. In particular, we compare users with 1) detailed goal instructions that conform to the system boundaries (closed-goal) and 2) vague goal instructions that are often unsupported but realistic (open-goal). Our study reveals that conversations in open-goal settings lead to catastrophic failures of the system, in which 92% of the dialogues had significant issues. Moreover, we conduct a thorough analysis to identify distinctive features between the two settings through error annotation. From this, we discover a novel “pretending” behavior, in which the system pretends to handle the user requests even though they are beyond the system’s capabilities. We discuss its characteristics and toxicity while showing recent large language models can also suffer from this behavior.</abstract>
<identifier type="citekey">kim-etal-2024-revealing</identifier>
<location>
<url>https://aclanthology.org/2024.nlp4convai-1.3</url>
</location>
<part>
<date>2024-08</date>
<extent unit="page">
<start>37</start>
<end>55</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Revealing User Familiarity Bias in Task-Oriented Dialogue via Interactive Evaluation
%A Kim, Takyoung
%A Shin, Jamin
%A Kim, Young-Ho
%A Bae, Sanghwan
%A Kim, Sungdong
%Y Nouri, Elnaz
%Y Rastogi, Abhinav
%Y Spithourakis, Georgios
%Y Liu, Bing
%Y Chen, Yun-Nung
%Y Li, Yu
%Y Albalak, Alon
%Y Wakaki, Hiromi
%Y Papangelis, Alexandros
%S Proceedings of the 6th Workshop on NLP for Conversational AI (NLP4ConvAI 2024)
%D 2024
%8 August
%I Association for Computational Linguistics
%C Bangkok, Thailand
%F kim-etal-2024-revealing
%X Most task-oriented dialogue (TOD) benchmarks assume users that know exactly how to use the system by constraining the user behaviors within the system’s capabilities via strict user goals, namely “user familiarity” bias. This data bias deepens when it combines with data-driven TOD systems, as it is impossible to fathom the effect of it with existing static evaluations. Hence, we conduct an interactive user study to unveil how vulnerable TOD systems are against realistic scenarios. In particular, we compare users with 1) detailed goal instructions that conform to the system boundaries (closed-goal) and 2) vague goal instructions that are often unsupported but realistic (open-goal). Our study reveals that conversations in open-goal settings lead to catastrophic failures of the system, in which 92% of the dialogues had significant issues. Moreover, we conduct a thorough analysis to identify distinctive features between the two settings through error annotation. From this, we discover a novel “pretending” behavior, in which the system pretends to handle the user requests even though they are beyond the system’s capabilities. We discuss its characteristics and toxicity while showing recent large language models can also suffer from this behavior.
%U https://aclanthology.org/2024.nlp4convai-1.3
%P 37-55
Markdown (Informal)
[Revealing User Familiarity Bias in Task-Oriented Dialogue via Interactive Evaluation](https://aclanthology.org/2024.nlp4convai-1.3) (Kim et al., NLP4ConvAI-WS 2024)
ACL