Proceedings of the 2023 ACM/IEEE International Conference on Human-Robot Interaction 2023
DOI: 10.1145/3568162.3576968
|View full text |Cite
|
Sign up to set email alerts
|

A Social Robot Reading Partner for Explorative Guidance

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2

Citation Types

0
2
0

Year Published

2023
2023
2024
2024

Publication Types

Select...
2
2

Relationship

0
4

Authors

Journals

citations
Cited by 4 publications
(2 citation statements)
references
References 22 publications
0
2
0
Order By: Relevance
“…We created an 11-element vector, called š‘  š‘” , to represent the observation state for the coaching dialogue flow, which consists of: prediction of interaction rupture (present or absent), current well-being exercise (savouring, gratitude, accomplishment, one door closes one door opens), speech features (duration of speech and silence), and previous actions (summarisation, follow-up question, and new episode). All of these features were collected at the end of each turn š‘” to keep track of the dialogue flow and the conversational interchange between the human coachee and the robotic coach, as in [52]. The actions š‘Ž š‘” were 3 discrete dialogue actions of the robotic coach that can decide the coaching dialogue flow of the well-being practice, namely (1) summarise what the coachee said, (2) ask for a follow-up question (e.g., "How does this event make you feel?…”
Section: Reinforcement Learning Problem Formulationmentioning
confidence: 99%
See 1 more Smart Citation
“…We created an 11-element vector, called š‘  š‘” , to represent the observation state for the coaching dialogue flow, which consists of: prediction of interaction rupture (present or absent), current well-being exercise (savouring, gratitude, accomplishment, one door closes one door opens), speech features (duration of speech and silence), and previous actions (summarisation, follow-up question, and new episode). All of these features were collected at the end of each turn š‘” to keep track of the dialogue flow and the conversational interchange between the human coachee and the robotic coach, as in [52]. The actions š‘Ž š‘” were 3 discrete dialogue actions of the robotic coach that can decide the coaching dialogue flow of the well-being practice, namely (1) summarise what the coachee said, (2) ask for a follow-up question (e.g., "How does this event make you feel?…”
Section: Reinforcement Learning Problem Formulationmentioning
confidence: 99%
“…The results showed that there were not significant differences among the three models. We then chose the DQN as it is the most commonly used one in discrete problems, e.g., [52].…”
Section: Reinforcement Learning Problem Formulationmentioning
confidence: 99%