Design and implementation of a user-adaptive website with information palettes

Abstract

The majority of existing websites on the Internet do not adapt to the individual user. Instead, they serve the same static content that has been created beforehand to everyone who visits the site. However, it has been shown that different people have different cognitive styles, or preferred ways in which they think, perceive information, and solve problems. Each cognitive style desires a certain type of information presented in a certain way. In this thesis, I design and implement a framework for creating user-adaptive websites that can infer a user's cognitive style from the webpages he or she visits and serve adaptive information palettes with content suited for that cognitive style.Specifically, the system first assigns ratings to each webpage, defining how each one rates along a set of cognitive style dimensions. Then it tracks a user's session on a website, compares it to sessions of past users, clusters similar sessions together, and computes the likely cognitive style of the user using a weighted average of the ratings of the webpages in the user's current session and in the cluster. I implemented this system as a customer advocacy website for General Motors. The website successfully infers users' cognitive styles and displays suitable information palettes.