PERSONALIZABLE DIRECTORIES: A CASE STUDY IN AUTOMATIC USER MODELLING

Abstract

A personalized directory system is presented which demonstrates how automatic user modelling can reduce the average number of menu-based selections needed to retrieve entries from a large ordered list, such as a telephone directory. It offers a simple case study for pilot human factor experiments in automatic user modelling; it affords the possibility of a device independent menu-selection system; in automatic user modelling are raised, such as the user's perception of a dynamic system, the need of user control and the measurement of user satisfaction. The personalized directory system is being used as a test-bed to examine such issues, and will refine our insight into the human factors implications of user modelling.