A Wizard-of-Oz Interface and Persona-based Methodology for Collecting Health Counseling Dialog

William R. Kearns, Neha Kaura, Myra Divina, Cuong Vo, Dong Si, Teresa Ward, Weichao Yuwen

Abstract: Health dialog collection is the primary bottleneck for the training and deployment of conversational agents into clinical practice. Current tools for the development of dialog systems are primarily focused on writing intent-slot schemas for natural language understanding and finite-state models of dialog management. However, this is a time-consuming process that can be opaque for clinical teams and the limitations of both approaches are well understood. The health domain presents additional barriers including access to patients and concerns about exposing personal health information. The contribution of this work is two-fold. First, it describes an interface designed to augment the ability of clinicians to efficiently engage in high-quality and empathetic health counseling dialog with their patients. Second, it presents a WOZ protocol for collecting this dialog using standardized patient actors each playing a role from a pool of caregiver personas.