View article

This chapter reports on research in the development of collaborative intelligent agents. This text argues that intelligent agents, both virtual and robotic, must know about more than just the task they are performing during collaboration with a human. They must also be capable of managing engagement with their human partners, expressing affect as it pertains to the collaboration, and modeling the relationship the agent develops with the user over time. The present chapter draws on examples of human-to-human collaborations to illustrate how affective expressions play a role in the outcome of the collaboration. It also presents numerous examples of research with agents that include engagement interaction during dialog and model relationships, while being collaborators with human users.