Designing and prototyping a "social" educational dialogue system using Wizard of Oz

Interpersonal closeness, or, rapport is an important element of human interaction, particularly in educational contexts, but current educational technologies are often unable to detect or respond to students in ways that draw on research on interpersonal social bonding in learning. In the "Rapport-Aligned Peer Tutor" (or, RAPT) project, we study rapport-building behaviors in peer tutoring in order to design a virtual peer tutor that can build rapport with students.

In this component of the project, we are designing and prototyping a virtual peer tutoring agent as an experimental platform to investigate theories of social dialogue in learning. We have developed a partially Wizard of Oz'ed dialogue system, informed by theories of interpersonal rapport between people, and partially driven by machine learning modules trained on annotated dialogue data from students learning collaboratively, and in partnership with other work from our lab on building "socially-aware" modules of a dialogue system for other applications. We have been conducted initial pilots of this system in a Wizard of Oz study.