Personality-based design and implementation for personalization of e-Commerce applications

Abstract

In commerce, user behaviour modelling is an important component of marketing and advertising. Personalization, a widely used feature of e-Commerce systems, is one aspect of such modelling. Current personalization systems require users to register in order to provide their services; and many of the personalization aspects offered are determined by requiring the user to fill extensive forms regarding their preferences. Although there are a few systems that provide automatic personalization, most focus on the link and content personalization only. Work done in user profiling and determining user models for marketing and commerce purposes have little emphasis on user personality; moreover, there has been a growing concern for privacy of information from the ever-growing online community. The objective of this thesis is to provide a model where the personality of the customer is used for personalization. A mechanism is proposed in which personalization does not require registration, the customer personality profile is determined implicitly, and the personalization module is updated dynamically while the customer browses the products. The design, implementation and testing of a bookstore using this model is presented; furthermore the system is coupled with a privacy component to present the concept of privacy-enhanced personalization of web pages.