Evaluating human-artificial agent decision congruence in a coordinated action task

Gaurav Patil; Phillip Bagala; Patrick Nalepka; Rachel W. Kallen; Michael J. Richardson

doi:10.1145/3527188.3563923

Evaluating human-artificial agent decision congruence in a coordinated action task

Gaurav Patil, Phillip Bagala, Patrick Nalepka, Rachel W. Kallen, Michael J. Richardson

School of Psychological Sciences

Research output: Chapter in Book/Report/Conference proceeding › Conference proceeding contribution › peer-review

1 Citation (Scopus)

Abstract

Recommender systems designed to augment human decision-making in multi-agent tasks need to not only recommend actions that align with the task goal, but which also maintain coordinative behaviors between agents. Further, if these systems are to be used for skill training, they need to impart implicit learning to its users. This work compared a recommender system trained using deep reinforcement learning to a heuristic-based system in recommending actions to human participants teaming with an artificial agent during a collaborative problem-solving task. In addition to evaluating task performance and learning, we also evaluate the extent to which the human action are congruent with the recommended actions.

Original language	English
Title of host publication	HAI 2022
Subtitle of host publication	Proceedings of the 10th Conference on Human-Agent Interaction
Place of Publication	New York, NY
Publisher	Association for Computing Machinery, Inc
Pages	327-329
Number of pages	3
ISBN (Electronic)	9781450393232
DOIs	https://doi.org/10.1145/3527188.3563923
Publication status	Published - 5 Dec 2022
Event	International Conference on Human-Agent Interaction (10th : 2022) - Christchurch, New Zealand Duration: 5 Dec 2022 → 8 Dec 2022

Conference

Conference	International Conference on Human-Agent Interaction (10th : 2022)
Abbreviated title	HAI ’22
Country/Territory	New Zealand
City	Christchurch
Period	5/12/22 → 8/12/22

Keywords

decision making
hierarchical deep reinforcement learning
multi-agent coordination
recommender system
shepherding

Access to Document

10.1145/3527188.3563923

Cite this

@inproceedings{fbafc585983047618571da9ed7ed67fc,

title = "Evaluating human-artificial agent decision congruence in a coordinated action task",

abstract = "Recommender systems designed to augment human decision-making in multi-agent tasks need to not only recommend actions that align with the task goal, but which also maintain coordinative behaviors between agents. Further, if these systems are to be used for skill training, they need to impart implicit learning to its users. This work compared a recommender system trained using deep reinforcement learning to a heuristic-based system in recommending actions to human participants teaming with an artificial agent during a collaborative problem-solving task. In addition to evaluating task performance and learning, we also evaluate the extent to which the human action are congruent with the recommended actions.",

keywords = "decision making, hierarchical deep reinforcement learning, multi-agent coordination, recommender system, shepherding",

author = "Gaurav Patil and Phillip Bagala and Patrick Nalepka and Kallen, {Rachel W.} and Richardson, {Michael J.}",

year = "2022",

month = dec,

day = "5",

doi = "10.1145/3527188.3563923",

language = "English",

pages = "327--329",

booktitle = "HAI 2022",

publisher = "Association for Computing Machinery, Inc",

note = "International Conference on Human-Agent Interaction (10th : 2022), HAI {\textquoteright}22 ; Conference date: 05-12-2022 Through 08-12-2022",

}

Patil, G, Bagala, P, Nalepka, P , Kallen, RW & Richardson, MJ 2022, Evaluating human-artificial agent decision congruence in a coordinated action task. in HAI 2022: Proceedings of the 10th Conference on Human-Agent Interaction. Association for Computing Machinery, Inc, New York, NY, pp. 327-329, International Conference on Human-Agent Interaction (10th : 2022), Christchurch, New Zealand, 5/12/22. https://doi.org/10.1145/3527188.3563923

Evaluating human-artificial agent decision congruence in a coordinated action task. / Patil, Gaurav; Bagala, Phillip; Nalepka, Patrick et al.
HAI 2022: Proceedings of the 10th Conference on Human-Agent Interaction. New York, NY: Association for Computing Machinery, Inc, 2022. p. 327-329.

Research output: Chapter in Book/Report/Conference proceeding › Conference proceeding contribution › peer-review

TY - GEN

T1 - Evaluating human-artificial agent decision congruence in a coordinated action task

AU - Patil, Gaurav

AU - Bagala, Phillip

AU - Nalepka, Patrick

AU - Kallen, Rachel W.

AU - Richardson, Michael J.

PY - 2022/12/5

Y1 - 2022/12/5

N2 - Recommender systems designed to augment human decision-making in multi-agent tasks need to not only recommend actions that align with the task goal, but which also maintain coordinative behaviors between agents. Further, if these systems are to be used for skill training, they need to impart implicit learning to its users. This work compared a recommender system trained using deep reinforcement learning to a heuristic-based system in recommending actions to human participants teaming with an artificial agent during a collaborative problem-solving task. In addition to evaluating task performance and learning, we also evaluate the extent to which the human action are congruent with the recommended actions.

AB - Recommender systems designed to augment human decision-making in multi-agent tasks need to not only recommend actions that align with the task goal, but which also maintain coordinative behaviors between agents. Further, if these systems are to be used for skill training, they need to impart implicit learning to its users. This work compared a recommender system trained using deep reinforcement learning to a heuristic-based system in recommending actions to human participants teaming with an artificial agent during a collaborative problem-solving task. In addition to evaluating task performance and learning, we also evaluate the extent to which the human action are congruent with the recommended actions.

KW - decision making

KW - hierarchical deep reinforcement learning

KW - multi-agent coordination

KW - recommender system

KW - shepherding

UR - http://www.scopus.com/inward/record.url?scp=85144608568&partnerID=8YFLogxK

UR - http://purl.org/au-research/grants/arc/FT180100447

U2 - 10.1145/3527188.3563923

DO - 10.1145/3527188.3563923

M3 - Conference proceeding contribution

AN - SCOPUS:85144608568

SP - 327

EP - 329

BT - HAI 2022

PB - Association for Computing Machinery, Inc

CY - New York, NY

T2 - International Conference on Human-Agent Interaction (10th : 2022)

Y2 - 5 December 2022 through 8 December 2022

ER -

Evaluating human-artificial agent decision congruence in a coordinated action task

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

ARC - Future Fellowships: Modelling Human Perceptual-Motor Interaction for Human-Machine Applications

Cite this

Evaluating human-artificial agent decision congruence in a coordinated action task

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Projects

ARC - Future Fellowships: Modelling Human Perceptual-Motor Interaction for Human-Machine Applications

Cite this