Argumentation-Based Dialogue Games for Shared Control in Human-Robot Systems

Abstract

Dialogue can support exchange of ideas and discussion of options as a means to enable shared decision making for human-robot collaboration. However, dialogue that supports dynamic, evidencebacked exchange of ideas is a major challenge for today's human-robot systems. The work presented here investigates the application of argumentation-based dialogue games as the means to facilitate flexible interaction, including unscripted changes in initiative. Two main contributions are provided in this paper. First, a methodology for implementing multiple types of argumentation-based dialogues for human-robot interaction is detailed. This includes explanation about which types of dialogues are appropriate given the beliefs of the participants and how multiple dialogues can occur simultaneously while maintaining a consistent set of beliefs for the participants. Second, a formal definition is presented for the Treasure Hunt Game (THG), a test environment that provides rich opportunities for experimentation in shared human-robot control, as well as motivating and engaging experiences for human subjects.

Analysis of empirical results on argumentation-based dialogue to support shared decision making in a human-robot team

“…The experiments analyzed here are based on two user studies conducted using the ArgHRI framework to play an adaptation of the Treasure Hunt Game (THG) [21] [7]. Our version of the THG involves a human and a robot player who work together as a team.…”

“…We have developed a software framework called ArgHRI [7] which employs computational argumentationbased dialogue [5], [6] to enable a robot to interact with a human and exchange ideas. The formal theory allows the robot to manage its beliefs in a structured way.…”

“…1. Cases for different types of dialogues (from [7]), based on the beliefs of the robot (R) and its beliefs about the human (H). b ∈ R means that the robot believes b.…”

“…This formal model prescribes a structure for defining protocols for utterances associated with each type of dialogue, essentially a list of legal moves that each participant in the dialogue game is allowed to select from. In related work, we have defined the protocols for each of the types of dialogue implemented in ArgHRI [13], [7], including state machines that control the possible sequences of locutions, from start state to termination state. One of the advantages of employing argumentation-based dialogue games is that it has been proven formally that the rules for each type of dialogue guarantee termination [6].…”

“…The argumentation-based dialogue theory underlying the framework is described elsewhere [7]. Experimental results from the study that focus on task performance statistics (such as how long it took for human-robot teams to make decisions, how long it took for robots to perform actions and how far robots traveled) are also reported elsewhere [8].…”