An overarching goal of Artificial Intelligence (AI) is creating autonomous, social agents that help people. Two important challenges, though, are that different people prefer different assistance from agents and that preferences can change over time. Thus, helping behaviors should be tailored to how an individual feels during the interaction. We hypothesize that human nonverbal behavior can give clues about users' preferences for an agent's helping behaviors, augmenting an agent's ability to computationally predict such preferences with machine learning models. To investigate our hypothesis, we collected data from 194 participants via an online survey in which participants were recorded while playing a multiplayer game. We evaluated whether the inclusion of nonverbal human signals, as well as additional context (e.g., via game or personality information), led to improved prediction of user preferences between agent behaviors compared to explicitly provided survey responses. Our results suggest that nonverbal communication -- a common type of human implicit feedback -- can aid in understanding how people want computational agents to interact with them.
more »
« less
Perceptions of the Helpfulness of Unexpected Agent Assistance
Much prior work on creating social agents that assist users relies on preconceived assumptions of what it means to be helpful. For example, it is common to assume that a helpful agent just assists with achieving a user’s objective. However, as assistive agents become more widespread, human-agent interactions may be more ad-hoc, providing opportunities for unexpected agent assistance. How would this affect human notions of an agent’s helpfulness? To investigate this question, we conducted an exploratory study (N=186) where participants interacted with agents displaying unexpected, assistive behaviors in a Space Invaders game and we studied factors that may influence perceived helpfulness in these interactions. Our results challenge the idea that human perceptions of the helpfulness of unexpected agent assistance can be derived from a universal, objective definition of help. Also, humans will reciprocate unexpected assistance, but might not always consider that they are in fact helping an agent. Based on our findings, we recommend considering personalization and adaptation when designing future assistive behaviors for prosocial agents that may try to help users in unexpected situations.
more »
« less
- Award ID(s):
- 2106690
- NSF-PAR ID:
- 10380039
- Date Published:
- Journal Name:
- Proceedings of 10th International Conference on Human-Agent Interaction (HAI)
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Prior work has shown that embodiment can benefit virtual agents, such as increasing rapport and conveying non-verbal information. However, it is unclear if users prefer an embodied to a speech-only agent for augmented reality (AR) headsets that are designed to assist users in completing real-world tasks. We conducted a study to examine users' perceptions and behaviors when interacting with virtual agents in AR. We asked 24 adults to wear the Microsoft HoloLens and find objects in a hidden object game while interacting with an agent that would offer assistance. We presented participants with four different agents: voice-only, non-human, full-size embodied, and a miniature embodied agent. Overall, users preferred the miniature embodied agent due to the novelty of his size and reduced uncanniness as opposed to the larger agent. From our results, we draw conclusions about how agent representation matters and derive guidelines on designing agents for AR headsets.more » « less
-
The presence of voice activated personal assistants (VAPAs) in people's homes rises each year [31]. Industry efforts are invested in making interactions with VAPAs more personal by leveraging information from messages and calendars, and by accessing user accounts for 3rd party services. However, the use of personal data becomes more complicated in interpersonal spaces, such as people's homes. Should a shared agent access the information of many users? If it does, how should it navigate issues of privacy and control? Designers currently lack guidelines to help them design appropriate agent behaviors. We used Speed Dating to explore inchoate social mores around agent actions within a home, including issues of proactivity, interpersonal conflict, and agent prevarication. Findings offer new insights on how more socially sophisticated agents might sense, make judgements about, and navigate social roles and individuals. We discuss how our findings might impact future research and future agent behaviors.more » « less
-
Kolawole, Olatunji Matthew (Ed.)more » « less
Persons living with human immunodeficiency virus (HIV) have a disproportionately higher burden of human papillomavirus infection (HPV)-related cancers. Causal factors include both behavioral and biological. While pharmaceutical and care support interventions help address biological risk of coinfection, as social conditions are common drivers of behaviors, structural interventions are key part of behavioral interventions. Our objective is to develop a joint HIV-HPV model to evaluate the contribution of each factor, to subsequently inform intervention analyses. While compartmental modeling is sufficient for faster spreading HPV, network modeling is suitable for slower spreading HIV. However, using network modeling for jointly modeling HIV and HPV can generate computational complexities given their vastly varying disease epidemiology and disease burden across sub-population groups. We applied a recently developed mixed agent-based compartmental (MAC) simulation technique, which simulates persons with at least one slower spreading disease and their immediate contacts as agents in a network, and all other persons including those with faster spreading diseases in a compartmental model, with an evolving contact network algorithm maintaining the dynamics between the two models. We simulated HIV and HPV in the U.S. among heterosexual female, heterosexual male, and men who have sex with men (men only and men and women) (MSM), sub-populations that mix but have varying HIV burden, and cervical cancer among women. We conducted numerical analyses to evaluate the contribution of behavioral and biological factors to risk of cervical cancer among women with HIV. The model outputs for HIV, HPV, and cervical cancer compared well with surveillance estimates. Model estimates for relative prevalence of HPV (1.67 times) and relative incidence of cervical cancer (3.6 times), among women with HIV compared to women without, were also similar to that reported in observational studies in the literature. The fraction attributed to biological factors ranged from 22–38% for increased HPV prevalence and 80% for increased cervical cancer incidence, the remaining attributed to behavioral. The attribution of both behavioral and biological factors to increased HPV prevalence and cervical cancer incidence suggest the need for behavioral, structural, and pharmaceutical interventions. Validity of model results related to both individual and joint disease metrics serves as proof-of-concept of the MAC simulation technique. Understanding the contribution of behavioral and biological factors of risk helps inform interventions. Future work can expand the model to simulate sexual and care behaviors as functions of social conditions to jointly evaluate behavioral, structural, and pharmaceutical interventions for HIV and cervical cancer prevention.
-
more » « less
Abstract Background Transfemoral prosthesis users’ high fall rate is related to increased injury risk, medical costs, and fear of falling. Better understanding how stumble conditions (e.g., participant age, prosthesis type, side tripped, and swing phase of perturbation) affect transfemoral prosthesis users could provide insight into response deficiencies and inform fall prevention interventions.
Methods Six unilateral transfemoral prosthesis users experienced obstacle perturbations to their sound limb in early, mid, and late swing phase. Fall outcome, recovery strategy, and kinematics of each response were recorded to characterize (1) recoveries versus falls for transfemoral prosthesis users and (2) prosthesis user recoveries versus healthy adult recoveries.
Results Out of 26 stumbles, 15 resulted in falls with five of six transfemoral prosthesis users falling at least once. By contrast, in a previously published study of seven healthy adults comprising 214 stumbles using the same experimental apparatus, no participants fell. The two oldest prosthesis users fell after every stumble, stumbles in mid swing resulted in the most falls, and prosthesis type was not related to strategy/fall outcomes. Prosthesis users who recovered used the elevating strategy in early swing, lowering strategy in late swing, and elevating or lowering/delayed lowering with hopping in mid swing, but exhibited increased contralateral (prosthetic-side) thigh abduction and trunk flexion relative to healthy controls. Falls occurred if the tripped (sound) limb did not reach ample thigh/knee flexion to sufficiently clear the obstacle in the elevating step, or if the prosthetic limb did not facilitate a successful step response after the initial sound-side elevating or lowering step. Such responses generally led to smaller step lengths, less anterior foot positioning, and more forward trunk flexion/flexion velocity in the resulting foot-strikes.
Conclusions Introducing training (e.g., muscle strength or task-specific motor skill) and/or modifying assistive devices (e.g., lower-limb prostheses or exoskeletons) may improve responses for transfemoral prosthesis users. Specifically, training or exoskeleton assistance could help facilitate sufficient thigh/knee flexion for elevating; training or prosthesis assistance could provide support-limb counteracting torques to aid in elevating; and training or prosthesis assistance could help initiate and safely complete prosthetic swing.
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
