Towards Modeling Confidentiality in Persuasive Robot Dialogue
Contribution to Book
International Conference on Smart Homes and Health Telematics ICOST 2016: Inclusive Smart Cities and Digital Health
In many persuasive health interventions, humanoid robots and other intelligent systems are capable of carrying out meaningful conversations with human subjects in an effort to influence humans towards behavior or attitudinal change. In human-to-human conversations, the listening party often has the ability to discern whether or not certain aspects of the conversation should be kept confidential. Consequently, in conversational service robot scenarios (including elderly care use cases), humans often have the expectation that humanoid robots are capable of preserving the privacy and confidentiality of a given human-robot dialogue. In this literature, we explore the inherent challenges and approaches to modeling confidentiality in human-robot interaction (HRI) dialogue scenarios involving a cloud-enabled networked robot. As a result, we share a novel reference model for designing persuasive dialogue systems.