Explores advanced language-based interfaces, both speech-only and multi-modal. Our research includes dialog management, natural language generation and synthesis. We are interested in the integration of different levels of processing and in the detection and robust handling of misunderstandings between humans and machines.