In this paper, we examine the application of electronic marketplaces, populated by buying and selling agents representing their human users, learning about potential business partners and making recommendations to their users. We propose a novel incentive mechanism to address the unfair rating problem arising when modeling the trustworthiness of selling agents relies on propagation of ratings provided by buying agents. In our mechanism, buying agents model other buyers and select the most trustworthy ones as their neighbors from which they can ask advice about sellers. In order to build reputation, sellers will model the reputation of buyers based on the number of their neighborhoods and increase values of products to satisfy reputable buyers. In consequence, our mechanism creates incentive for buyers to provide fair ratings of sellers. We also discuss how a marketplace designed in this way leads to better profit both for buyers and sellers and as such fosters trust between the agents and their human users.