Architecture of a dialog system with an assistant agent

This paper describes a dialog system allowing a human user and a personal assistant agent to communicate through natural language. The Assistant uses explicit task representation and knowledge representation to achieve an intelligent dialog. The dialog architecture is independent from the knowledge or task representation and includes different models: dialog models, task models, user models and domain models. The paper describes how the dialog manager uses the models to conduct the dialog and how the approach is related to other approaches in the same domain. The system has been implemented in a collaborative environment in order to personalize services and to improve interaction.