Design of Evolutionary Product-Brokering Agents for E-Commerce

S.-U. Guan, C.S. Ngoo, and F. Zhu (Singapore)


productbrokering agent, ecommerce, evolution, profiling, fitness function, genetic algorithms.


One of the potential uses of an agent-based system has been in the area of e-commerce. A lot of research has been done on making the system intelligent enough to personalize its services for the user. User-supplied are normally used to generate a profile of the user. In this paper, a design for an evolutionary ontology-based product-brokering agent has been proposed. It uses an evaluation function to represent the user’s preference instead of the usual keyword-based profile. By using genetic algorithms, the agent tries to track the user’s preferences for a particular product by tuning some of the parameters inside this function. A prototype has been implemented in Java and the results obtained from experiments looks promising.

Important Links:

Go Back