Master of Web Science

Abstract

As the web has spread widely in the past years, it is now very important to personalize and customize websites so that we can improve user experience. Customer profiling is one of the most important methods to add personalization and customization to websites as it can capture the properties and interests of the customer in order to use them later to service him properly. In this paper, we present a generalized automatic customer profiling model that can suite most websites. Our model is based on probability theory; it captures user behavior slots in order to predict his profile. These probabilistic slots values are used to generate the customer adequate services. We show how to incorporate this model in a web service so that any website developer can use this service to add automatic user profiling to his website.