IEEE Intelligent Informatics Bulletin
One of the goals of the field of human-robot interaction is to enable robots to interact through natural language. This is a particularly challenging problem due to the uncertain and open nature of most application domains. In this paper, we summarize our recent work in developing natural language understanding and generation algorithms. These algorithms are specifically designed to handle the uncertain and open-worlds in which interactive robots must operate, and use a Consultant Framework specifically designed to account for the realities of integrated robot architectures.