Adaptive user modelling in an intelligent telephone assistant

Abstract

With the burgeoning complexity and capabilities of moderninformation appliances and services, user modelling is becoming anincreasingly important research area. Simple user profiles already personalisemany software products and consumer goods such as digital TV recorders andmobile phones. A user model should be easy to initialise, and it must adapt inthe light of interaction with the user. In many cases, a large amount of trainingdata is needed to generate a user model, and adaptation is equivalent toretraining the system. This paper briefly outlines the user modelling problemand work done at BTexact on an Intelligent Personal Assistant (IPA) whichincorporates a user profile. We go on to describe FILUM, a more flexiblemethod of user modelling, and show its application to the Telephone Assistantcomponent of the IPA, with tests to illustrate its usefulness.