We describe Scusi?, a multi-stage, spoken language interpretation mechanism designed to be part of a robot-mounted dialogue system. Scusi?'s interpretation process maps spoken utterances to conceptual graphs, and the nodes in these graphs to concepts in the world. Maximum posterior probability is used to rank the (partial) interpretations produced at each stage of this process.
The paper is available as pdf (120 kB).
Also available on the Springer-Verlag website