Intelligent agents for an artificial market system

Publisher Website
Google Scholar
Abstract
This paper describes an agent- based artificial market system whose underlying interaction protocols provide advanced features. Using the system, actors (i. e., customers and merchants) can delegate a variety of tasks to personal intelligent agents that act as their artificial employees. Contrary to other approaches, where a new agent is launched when their associated actors intend to perform a buying or selling transaction and “lives” only while this transaction is processed, our approach builds on a personalization of agents that permanently “live” in the market representing their actors' interests. Beyond just requesting and proposing an offer, agents in our system maintain a profile of their owners, which is updated upon the actor- agent interaction type. Furthermore, they can proactively ask their owners permission to initiate a transaction (e. g., when a new product, which match ones profile, appears in the market). The system is also enabled with a highly interactive multiple criteria decision making tool that can handle ill- structured information during a purchase transaction, and perform a progressive synthesis and comparative evaluation of the existing proposals.

This publication has 7 references indexed in Scilit: