A bidding mechanism for Web-based agents involved in information classification

The impact of the World Wide Web on providing an easy information access is clearly evident in all aspects of today's life. As attractive as the information availability is, due to its sheer volume, it creates an information-overload on users. Agent-based collaborative filtering is a technique used to effectively counter this burden. For agents to collaborate successfully, and maintain an overall progress of the entire interconnected environment, a governing mechanism is necessary. In this article, we present an economic framework based on a bidding mechanism for agents to communicate and negotiate with each other, thereby achieving collaborative information classification. D-SIFTER, a system developed using this economic framework, is described along with various experiments and their results.