Learning users’ assistance requirements with WATSON

Abstract

Interface agents are computer programs that provide personalized assistance to users with their computerbased tasks. The interface agents developed so far have focused their attention on learning a user's preferences in a given application domain and on assisting him according to them. However, in order to personalize the interaction with users, interface agents should also learn how to best interact with each user and how to provide them assistance of the right sort at the right time. Particularly, an interface agent has to determine when the user wants a suggestion to solve a problem, when he requires only a warning about it, when he wants the agent to execute an action to deal with the problem and when he wants the agent to do just nothing. In this work we propose a learning algorithm, named WATSON, to tackle this problem. The WATSON algorithm enables an interface agent to adapt its behavior and its interaction with a user to the user's assistance requirements.