We describe an architecture for spoken dialogue interfaces to semi-autonomoussystems that transforms speech signals through successive representations oflinguistic, dialogue, and domain knowledge. Each step produces an output, and ameta-output describing the transformation, with an executable program in asimple scripting language as the final result. The output/meta-outputdistinction permits perspicuous treatment of diverse tasks such as resolvingpronouns, correcting user misconceptions, and optimizing scripts.
展开▼
