A Multi-Tiered Recommender System Architecture for Supporting E-Commerce

Nowadays, many e-Commerce tools support customers with automatic recommendations. Many of them are centralized and lack in efficiency and scalability, while other ones are distributed and require a computational overhead excessive for many devices. Moreover, all the past proposals are not ``open'' and do not allow new personalized terms to be introduced into the domain ontology. In this paper, we present a distributed recommender, based on a multi-tiered agent system, trying to face the issues outlined above. The proposed system is able to generate very effective suggestions without a too onerous computational task. We show that our system introduces significant advantages in terms of openess, privacy and security.