Economic Recommendation with Surplus Maximization

The World Wide Web makes it easy for users to access various services online, such as online shopping in e-commerce, online financing in P2P lending networks, and even working online on freelancing platforms. A fundamentally important function of web applications is Online Service Allocation (OSA), which means matching service consumers with ‘appropriate’ goods (i.e., products/loans/jobs) from service producers. Since consumers often have the freedom to choose goods, OSA function is normally provided in the form of personalized recommendation or search. Existing search and recommendation system metrics either measure the benefit of individual consumers or the profit of individual producers. However, a sustainable web platform needs to be beneficial for both consumers and producers, otherwise one party might stop using it. In this paper, we aim to promote the web intelligence for social good. We show how to adapt economists’ traditional idea of maximizing total surplus (the sum of consumer net benefit and producer profit) to the heterogeneous world of online service allocation (OSA). Modifications of traditional personalized recommendation algorithms enable us to apply total surplus maximization (TSM) to three very different types of real-world tasks (e-commerce, P2P lending, and freelancing). Our experimental results on three tasks suggest that TSM compares very favorably to currently popular approaches, to the benefit of both producers and consumers.