The perception of emotion in artificial agents

Hortensius, Ruud; Hekele, Felix; Cross, Emily S.

doi:10.31234/osf.io/ufz5w

Cited by 9 publications

(10 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Sentiment analysis is also known as opinion or emotion artificial intelligence [ 6 ]. It refers to the use of natural language processing (NLP), text analysis, computational linguistics, and machine learning to systematically analyze people’s written language [ 7 ].…”

Section: Introductionmentioning

confidence: 99%

Emotion diffusion effect: Negative sentiment COVID-19 tweets of public organizations attract more responses from followers

Yang

et al. 2022

PLoS ONE

View full text Add to dashboard Cite

Coronavirus disease 2019 (COVID-19) has triggered an enormous number of discussion topics on social media Twitter. It has an impact on the global health system and citizen responses to the pandemic. Multiple responses (replies, favorites, and retweets) reflect the followers’ attitudes and emotions towards these tweets. Twitter data such as these have inspired substantial research interest in sentiment and social trend analyses. To date, studies on Twitter data have focused on the associational relationships between variables in a population. There is a need for further discovery of causality, such as the influence of sentiment polarity of tweet response on further discussion topics. These topics often reflect the human perception of COVID-19. This study addresses this exact topic. It aims to develop a new method to unveil the causal relationships between the sentiment polarity and responses in social media data. We employed sentiment polarity, i.e., positive or negative sentiment, as the treatment variable in this quasi-experimental study. The data is the tweets posted by nine authoritative public organizations in four countries and the World Health Organization from December 1, 2019, to May 10, 2020. Employing the inverse probability weighting model, we identified the treatment effect of sentiment polarity on the multiple responses of tweets. The topics with negative sentiment polarity on COVID-19 attracted significantly more replies (69±49) and favorites (688±677) than the positive tweets. However, no significant difference in the number of retweets was found between the negative and positive tweets. This study contributes a new method for social media analysis. It generates new insight into the influence of sentiment polarity of tweets about COVID-19 on tweet responses.

show abstract

Section: Introductionmentioning

confidence: 99%

Emotion diffusion effect: Negative sentiment COVID-19 tweets of public organizations attract more responses from followers

Yang

et al. 2022

PLoS ONE

View full text Add to dashboard Cite

show abstract

“…Although this explanation is not very likely, since the same procedure was used in the study by Van Nieuwenburg et al [6], future studies should consider to directly scenting the robot instead of using 'external' fragrances. Finally, cross-platform generalisability [48,49] could play a role. In a recent evaluation of trust questionnaires most often employed in human-robot interaction, issues of generalisability have been reported [32].…”

Section: Discussionmentioning

confidence: 99%

No effect of the smell of hexanal on trust in human-robot interaction

Croijmans¹,

Erp²,

Bakker³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

The level of interpersonal trust among people is partially determined through the sense of smell. Hexanal, a molecule which smell resembles freshly cut grass, can increase trust in people. Here, we ask the question if smell can be leveraged to facilitate human-robot interaction and test whether hexanal also increases the level of trust during collaboration with a social robot. In a preregistered double-blind, placebo-controlled study, we tested if trial-by-trial and general trust during perceptual decision making in collaboration with a social robot is affected by hexanal across two samples (n = 46 and n = 44). It was hypothesized that unmasked hexanal and hexanal masked by eugenol, a molecule with a smell resembling clove, would increase the level of trust in human-robot interaction, compared to eugenol alone or a control condition consisting of only the neutral smelling solvent propylene glycol. Contrasting previous findings in human interaction, no significant effect of unmasked or eugenol-masked hexanal on trust in robots was observed. These findings indicate that the conscious or nonconscious impact of smell on trust might not generalise to interactions with social robots. One explanation could be category- and context-dependency of smell leading to a mismatch between the natural smell of hexanal, a smell also occurring in human sweat, and the mechanical physical or mental representation of the robot.

show abstract

“…In contrast to Fraune ( 2020 ), we found that people who believed they played against robots showed an even stronger bias towards humanoid robots than humans. As these results are exploratory in nature and the explicit functional role of a robot’s form and shape in the perception and reaction to these agents remain to be understood ( Hortensius, Hekele & Cross, 2018 ), future studies could explore the effect of belief on social cognitive reactions to humanoid and mechanoid robots. Besides the importance of a robot’s appearance, other factors related to social, cognitive, and emotional capabilities of the robot might influence collaboration with a robot.…”

Section: Discussionmentioning

confidence: 99%

“…The purpose of Experiments 3 and 4 was to investigate whether individuals show intergroup empathy biases towards robots when they team up with one. We also examined the extent to which our findings might generalise across different robots, an unaddressed issue in the field of social robots ( Cross et al, 2019 ; Henschel et al, 2020 ; Hortensius et al, 2018 ). In Experiment 3, we used the Cozmo robot, a small entertainment robot with a mechanical design, while we used the NAO robot, a humanoid robot often used in HRI research in Experiment 4.…”

Section: Experiments 3 Andmentioning

confidence: 99%

Empathy and Schadenfreude in Human–Robot Teams

Jong

Hortensius

Hsieh

et al. 2021

Journal of Cognition

Self Cite

View full text Add to dashboard Cite

Intergroup dynamics shape the ways in which we interact with other people. We feel more empathy towards ingroup members compared to outgroup members, and can even feel pleasure when an outgroup member experiences misfortune, known as schadenfreude. Here, we test the extent to which these intergroup biases emerge during interactions with robots. We measured trial-by-trial fluctuations in emotional reactivity to the outcome of a competitive reaction time game to assess both empathy and schadenfreude in arbitrary human-human and human-robot teams. Across four experiments (total n = 361), we observed a consistent empathy and schadenfreude bias driven by team membership. People felt more empathy towards ingroup members than outgroup members and more schadenfreude towards outgroup members. The existence of an intergroup bias did not depend on the nature of the agent: the same effects were observed for human-human and human-robot teams. People reported similar levels of empathy and schadenfreude towards a human and robot player. The human likeness of the robot did not consistently influence this intergroup bias. In other words, similar empathy and schadenfreude biases were observed for both humanoid and mechanoid robots. For all teams, this bias was influenced by the level of team identification; individuals who identified more with their team showed stronger intergroup empathy and schadenfreude bias. Together, we show that similar intergroup dynamics that shape our interactions with people can also shape interactions with robots. Our results highlight the importance of taking intergroup biases into account when examining social dynamics of human-robot interactions.

show abstract

The perception of emotion in artificial agents

Cited by 9 publications

References 10 publications

Emotion diffusion effect: Negative sentiment COVID-19 tweets of public organizations attract more responses from followers

Emotion diffusion effect: Negative sentiment COVID-19 tweets of public organizations attract more responses from followers

No effect of the smell of hexanal on trust in human-robot interaction

Empathy and Schadenfreude in Human–Robot Teams

Contact Info

Product

Resources

About