Improving agent control for user modeling

This paper describes the User Modeling Agent (UM Agent) of an intelligent graphical user interface that manipulates files. The intelligent GUI is called IFM and it monitors users while they work; in case a user has made a mistake, it intervenes automatically and offers advice. The reasoning of the UM Agent is largely based on an adaptation of a cognitive theory, called Human Plausible Reasoning (HPR). The UM Agent observes users during their interaction with the system, maintains and manages the user profiles and provides relevant information whenever other agents request it. The main focus of this paper is on examining how the UM Agent´s control can be improved by the combination of Human Plausible Reasoning and stereotypical knowledge.