An intelligent brokering system to support multi-agent Web-based 4/sup th/-party logistics

Abstract

An intelligent agent-based framework that supports fourth-party logistics (4PL) operations on the Web is proposed. In our system, customers specify job requests over the Web dynamically. An e-marketplace allows intelligent third-party logistics (3PL) agents to bid for customers' job requests. The intelligence lies in the e-marketplace optimally deciding which agents' bids should be satisfied based on a set of predetermined factors (pricing, preferences and fairness). We model the underlying brokering problem as a set packing problem (SPP), an NP-hard optimization problem. An iterative greedy approximation algorithm is proposed to solve the SPP, and experimental results show its effectiveness against the classical greedy method proposed by Chvatal (1979).