School of Computing


Persuasive Dialogue

Since face-to-face communication is a highly efficient means of interaction, embodied conversation agents (virtual humans with speech and natural language capabilities) are a promising technology by which to realise user dialogues with computers. Indeed Reeves and Nass demonstrated that people willingly interpret artefacts as social actors, even where they display human-like characteristics. Previous studies of embodied conversation agents have primarily originated in the computer science community and as such have tended to be investigations of specialist applications. A key shortcoming of previous studies is the functionality of the application and the embedded agents themselves. Few actual embodied conversation agents have significant natural language processing capabilities, and character behaviour is typically canned or highly constrained. In this project we are investigating novel methods simulate both the physical and the cognitive capabilities of the ECA, allowing us to evaluate the persuasive potential of a highly functional ECA.

 Collaborators: Marco de Boni (Unilever Corporate Research)