AN INTELLIGENT MAN-MACHINE DIALOGUE SYSTEM BASED ON AI PLANNING

We describe the modular architecture of a generic dialogue system that assists a user/operator in performing a task with a tool. This coachi ng system is named CALLIOPE after the Greek goddess of eloquence. It a ims at being an active partner in an intelligent man-machine dialogue. The intelligent dimension of the coaching system is reflected by its ability to adapt to the user and the situation at hand. The CALLIOPE s ystem contains an explicit user model and world model to situate its d ialogue actions. A plan library allows it to follow loosely predetermi ned dialogue scenarios. The heart of the coaching system is an Al plan ning module, which plans a series of dialogue actions. We present a co herent set of three dialogue or speech actions that will make up the p hysical form of the man-machine communication. The use of the AI plann ing paradigm as a basis for man-machine interaction is motivated by re search in various disciplines, as e.g., AI, Cognitive Science and Soci al Sciences. Starting from the man-man communication metaphor, we can view the ''thinking before speaking'' of a human communication partner as constructing an underlying plan which is responsible for the purpo siveness, the organisation and the relevance of the communication. CAL LIOPE has been fully implemented and tested on theoretical examples. A t present, also three tailored versions of CALLIOPE are in operational use in different industrial application domains: operator support for remedying-tasks in chemical process industry, operator support for a combined task of planning, plan execution and process control in the a rea of chemical process development, and thirdly decision support in p roduction scheduling.