Towards evolution of software agents in electronic commerce

Abstract

With the development of Internet computing and software agent technologies, agent-based electronic commerce (e-commerce) is emerging. Software agents have demonstrated tremendous potential in conducting various tasks in e-commerce. However, when agents are initially created, they have little knowledge and experience with relatively lower capability. They should also strive to adapt themselves to the changing environment. It is advantageous if they have the ability to learn and evolve. This paper addresses evolution of software agents in e-commerce. Agent fitness and life cycle are proposed as evolution mechanisms, and modularized agent structure is introduced to facilitate the evolution process. Genetic Programming (GP) operators are employed to restructure agents in the proposed multi-agent evolution cycle.