A 2-phase evaluation model for agent-mediated Internet marketplaces

We propose a 2-phase evaluation model that can be applied to agent-mediated Internet marketplaces. The model not only facilitates parallel processing, it also allows mobile agents to select online shops (e-shops) and evaluate offers autonomously and automatically. To provide the "best" answer, the approach considers multiple attributes of an e-shop and an offer, such as the commercial credit and security rank of the e-shop, price, delivery/shipment service and warranty service of an offer. Moreover, with the fuzzy-set rules, the evaluation results are more reasonable and accord with human behavior. With the 2-phase evaluation of this model, it can not only help to control the scale of the dispatched mobile agent to a rational level to reduce the burden of network traffic, but also help to evaluate offers to assist users to make good decisions.