Socially Assistive Robotics (SAR) is the field of study dedicated to developing algorithms and models that enable robots to assist users in goal-directed interactions, often in situations connected to improving health and wellness or enriching education. In this work, we address one extension of SAR into multi-party interactions: robots acting as moderators, facilitating group interactions. We impose a computational structure on this problem by defining moderation as the problem of assigning resources to users in an interaction in order to achieve some pre-defined goal or goals, related to either the task (what the group is trying to achieve, such as assembling widgets) or to social aspects of the interaction (how the group should do the task; for example, maximizing the number of intra-group interactions). These resources can include both social resources such as the conversational floor or participant attention, and physical resources such as task-related objects. The approach is summarized in Figure 1. We provide an overview of a validation study of robot moderators in a group interaction, and describe ongoing work on extending this approach to moderation using symbolic planning.
[1]
Tetsunori Kobayashi,et al.
Four-participant group conversation: A facilitation robot controlling engagement density as the fourth participant
,
2015,
Comput. Speech Lang..
[2]
Ted Selker,et al.
Considerate Audio MEdiating Oracle (CAMEO): improving human-to-human communications in conference calls
,
2012,
DIS '12.
[3]
Yoshinori Kobayashi,et al.
A techno-sociological solution for designing a museum guide robot: Regarding choosing an appropriate visitor
,
2012,
2012 7th ACM/IEEE International Conference on Human-Robot Interaction (HRI).
[4]
Eric Horvitz,et al.
Facilitating multiparty dialog with gaze, gesture, and speech
,
2010,
ICMI-MLMI '10.
[5]
Maria Pateraki,et al.
Two people walk into a bar: dynamic multi-party social interaction with a robot agent
,
2012,
ICMI '12.