User interface optimization through breadth of distribution analysis

Abstract

A Java tutorial was developed as a World Wide Web (WWW) site for use in capturing user behavior data. Breadth of distribution analysis was then applied to the data collected in order to characterize the usage of the user interface through the shape, connectedness, and order of traversal of each user in the sample. The results reveal distinct user groups with different levels of user knowledge and needs in relation to the web site content The resulting user interface analysis process produces a set of recommendations for optimizing the user interface, including adaptive interfaces for different user subgroups and optimization of sequential rather than topical presentations