Web-based User Profiling using a Recurrent Neural Network

R. MacDonald and D.L. Silver (Canada)


User profiling, adaptive interface, artificial neural networks, ecommerce, clickstream analysis


The Internet based electronic market has grown so rapidly that companies must now strive to not only have a presence on the Internet but to differentiate themselves from their competitors. Personalization of the web shopping experience is one approach that E-Commerce companies can take to attract and retain customers. This paper describes a collaborative user profiling and adaptive interface method embedded within a fictitious E commerce website. The system combines JavaScript, Java Applet and Artificial Neural Network (ANN) technologies to produce web pages with dynamic content that recommend the next best links for a customer based on the interactions of prior customers. Tests with human subjects show (1) the effectiveness of the system when compared to a base-line system that randomly recommends best links and (2) the advantage of the system to users who were not informed of the purpose of the recommendation links.

Important Links:

Go Back