An extensible agent architecture for a competitive market-based allocation of consumer attention space

A competitive distributed recommendation mechanism is introduced based on adaptive software agents for efficiently allocating the ``customer attention space'', or banners. In the example of an electronic shopping mall, the task of correctly profiling and analyzing the customers is delegated to the individual shops that operate in a distributed, remote fashion. The evaluation and classification of customers for the bidding on banners is not handled by a central agency as is customary, but is a distributed process where all shops bidding for a customer partake. This allows each agent for a shop to apply its own private strategy, learning-mechanism, and specific domain knowledge without revealing sensitive business operations to the central party. We present a scalable and extensible software agent architecture and prototype for distributed market-based allocation of customer attention space. The agents can operate in multiple markets concurrently. The protocol for communication between the agents is designed for optimal performance of the system.