Computational argumentation to support multi-party human-robot interaction:Challenges and advantages
Elizabeth A Black, Elizabeth I Sklar · Research Portal (King's College London) · 2016
Recent work has explored the use of computational argumentation-based dialogues as a mechanism to support two-party human-robot interaction and collaborative decision making. Many real-world application scenarios involve interaction between multiple humans and multiple robots—not just two parties, i.e., not just one human and one robot. In this paper, a range of social issues are considered that arise in multiparty human-robot situations. These include issues of trust, privacy and ethics with respect to sharing information and modeling the beliefs of others. Using computational argumentation and argumentation-based dialogues can provide a sound basis for addressing the mitigating circumstances presented by these issues, and for reasoning under the uncertainties that these issues present. Here, a running example is developed, along with an approach for addressing the issues that is based on the application of computational argumentation strategies. A set of challenges are identified, and some directions for future research are proposed.