Artificial Immune System-Based Customer Data Clustering in an e-Shopping Application

摘要

We address the problem of adaptivity of the interaction between an e-shopping application and its users. Our approach is novel, as it is based on the construction of an Artificial Immune Network (AIN) in which a mutation process is incorporated and applied to the customer profile feature vectors. We find that the AIN-based algorithm yields clustering results of users' interests that are qualitatively and quantitatively better than the results achieved by using other more conventional clustering algorithms. This is demonstrated on user data that we collected using Vision. Com, an electronic video store application that we have developed as a test-bed for research purposes.